Upload folder using huggingface_hub
Browse filesThis view is limited to 50 files because it contains too many changes.
See raw diff
- checkpoint-27500/config.json +25 -0
- checkpoint-27500/generation_config.json +6 -0
- checkpoint-27500/merges.txt +0 -0
- checkpoint-27500/model.safetensors +3 -0
- checkpoint-27500/optimizer.pt +3 -0
- checkpoint-27500/rng_state_0.pth +3 -0
- checkpoint-27500/rng_state_1.pth +3 -0
- checkpoint-27500/scheduler.pt +3 -0
- checkpoint-27500/special_tokens_map.json +24 -0
- checkpoint-27500/tokenizer_config.json +22 -0
- checkpoint-27500/trainer_state.json +1958 -0
- checkpoint-27500/training_args.bin +3 -0
- checkpoint-27500/vocab.json +0 -0
- checkpoint-28000/config.json +25 -0
- checkpoint-28000/generation_config.json +6 -0
- checkpoint-28000/merges.txt +0 -0
- checkpoint-28000/model.safetensors +3 -0
- checkpoint-28000/optimizer.pt +3 -0
- checkpoint-28000/rng_state_0.pth +3 -0
- checkpoint-28000/rng_state_1.pth +3 -0
- checkpoint-28000/scheduler.pt +3 -0
- checkpoint-28000/special_tokens_map.json +24 -0
- checkpoint-28000/tokenizer_config.json +22 -0
- checkpoint-28000/trainer_state.json +1993 -0
- checkpoint-28000/training_args.bin +3 -0
- checkpoint-28000/vocab.json +0 -0
- checkpoint-28500/config.json +25 -0
- checkpoint-28500/generation_config.json +6 -0
- checkpoint-28500/merges.txt +0 -0
- checkpoint-28500/model.safetensors +3 -0
- checkpoint-28500/optimizer.pt +3 -0
- checkpoint-28500/rng_state_0.pth +3 -0
- checkpoint-28500/rng_state_1.pth +3 -0
- checkpoint-28500/scheduler.pt +3 -0
- checkpoint-28500/special_tokens_map.json +24 -0
- checkpoint-28500/tokenizer_config.json +22 -0
- checkpoint-28500/trainer_state.json +2028 -0
- checkpoint-28500/training_args.bin +3 -0
- checkpoint-28500/vocab.json +0 -0
- checkpoint-29000/config.json +25 -0
- checkpoint-29000/generation_config.json +6 -0
- checkpoint-29000/merges.txt +0 -0
- checkpoint-29000/model.safetensors +3 -0
- checkpoint-29000/optimizer.pt +3 -0
- checkpoint-29000/rng_state_0.pth +3 -0
- checkpoint-29000/rng_state_1.pth +3 -0
- checkpoint-29000/scheduler.pt +3 -0
- checkpoint-29000/special_tokens_map.json +24 -0
- checkpoint-29000/tokenizer_config.json +22 -0
- checkpoint-29000/trainer_state.json +2063 -0
checkpoint-27500/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_function": "gelu_new",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPTJForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attn_pdrop": 0.1,
|
| 7 |
+
"bos_token_id": 50256,
|
| 8 |
+
"embd_pdrop": 0.1,
|
| 9 |
+
"eos_token_id": 50256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"layer_norm_epsilon": 1e-05,
|
| 12 |
+
"model_type": "gptj",
|
| 13 |
+
"n_embd": 768,
|
| 14 |
+
"n_head": 12,
|
| 15 |
+
"n_inner": null,
|
| 16 |
+
"n_layer": 12,
|
| 17 |
+
"n_positions": 1024,
|
| 18 |
+
"resid_pdrop": 0.1,
|
| 19 |
+
"rotary_dim": 64,
|
| 20 |
+
"tie_word_embeddings": false,
|
| 21 |
+
"torch_dtype": "float32",
|
| 22 |
+
"transformers_version": "4.46.2",
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 50257
|
| 25 |
+
}
|
checkpoint-27500/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.46.2"
|
| 6 |
+
}
|
checkpoint-27500/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-27500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a7203fb54a4a505d0a7dcebc62b9ff2ac8aebf6ffedef31556799d53fbfa7eab
|
| 3 |
+
size 648995876
|
checkpoint-27500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8090e75d523b73a3eade256d2068eafbadd2f45d4d5d6eb6e10378187737f527
|
| 3 |
+
size 1298039738
|
checkpoint-27500/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8990c555e16a43a60226cfabe75274d56952a65e337b31153ebb3e4f97847bee
|
| 3 |
+
size 14512
|
checkpoint-27500/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:76620cf8bab91875184db637621c428d15520f5bdcc3ad9b24f50a7ec23cdf19
|
| 3 |
+
size 14512
|
checkpoint-27500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:99649e8bfadd84b609ed49defc9464b2679aeda73d702dcc5d7f9e63143e84f3
|
| 3 |
+
size 1000
|
checkpoint-27500/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|endoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endoftext|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "<|endoftext|>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<|endoftext|>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": true,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-27500/tokenizer_config.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"50256": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": true,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"bos_token": "<|endoftext|>",
|
| 15 |
+
"clean_up_tokenization_spaces": false,
|
| 16 |
+
"eos_token": "<|endoftext|>",
|
| 17 |
+
"errors": "replace",
|
| 18 |
+
"model_max_length": 1024,
|
| 19 |
+
"pad_token": "<|endoftext|>",
|
| 20 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 21 |
+
"unk_token": "<|endoftext|>"
|
| 22 |
+
}
|
checkpoint-27500/trainer_state.json
ADDED
|
@@ -0,0 +1,1958 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 187.7133105802048,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 27500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.6825938566552902,
|
| 13 |
+
"grad_norm": 1.5864567756652832,
|
| 14 |
+
"learning_rate": 0.0001,
|
| 15 |
+
"loss": 5.292,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.36518771331058,
|
| 20 |
+
"grad_norm": 0.6435245275497437,
|
| 21 |
+
"learning_rate": 0.0002,
|
| 22 |
+
"loss": 1.3885,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 2.04778156996587,
|
| 27 |
+
"grad_norm": 0.6205620169639587,
|
| 28 |
+
"learning_rate": 0.0003,
|
| 29 |
+
"loss": 1.2326,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 2.73037542662116,
|
| 34 |
+
"grad_norm": 0.668118953704834,
|
| 35 |
+
"learning_rate": 0.0004,
|
| 36 |
+
"loss": 1.234,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 3.4129692832764507,
|
| 41 |
+
"grad_norm": 0.5799328684806824,
|
| 42 |
+
"learning_rate": 0.0005,
|
| 43 |
+
"loss": 1.2338,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 4.09556313993174,
|
| 48 |
+
"grad_norm": 0.5505828261375427,
|
| 49 |
+
"learning_rate": 0.0006,
|
| 50 |
+
"loss": 1.2287,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 4.778156996587031,
|
| 55 |
+
"grad_norm": 0.515738844871521,
|
| 56 |
+
"learning_rate": 0.0007,
|
| 57 |
+
"loss": 1.2327,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 5.460750853242321,
|
| 62 |
+
"grad_norm": 0.509695291519165,
|
| 63 |
+
"learning_rate": 0.0008,
|
| 64 |
+
"loss": 1.2565,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 6.143344709897611,
|
| 69 |
+
"grad_norm": 0.6063594818115234,
|
| 70 |
+
"learning_rate": 0.0009000000000000001,
|
| 71 |
+
"loss": 1.2289,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 6.825938566552901,
|
| 76 |
+
"grad_norm": 0.5587486624717712,
|
| 77 |
+
"learning_rate": 0.001,
|
| 78 |
+
"loss": 1.2334,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 7.508532423208191,
|
| 83 |
+
"grad_norm": 0.5273250341415405,
|
| 84 |
+
"learning_rate": 0.0009999721746341338,
|
| 85 |
+
"loss": 1.2316,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 8.19112627986348,
|
| 90 |
+
"grad_norm": 0.593492329120636,
|
| 91 |
+
"learning_rate": 0.000999888701977651,
|
| 92 |
+
"loss": 1.2279,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 8.873720136518772,
|
| 97 |
+
"grad_norm": 0.44249382615089417,
|
| 98 |
+
"learning_rate": 0.000999749592353472,
|
| 99 |
+
"loss": 1.231,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 9.556313993174061,
|
| 104 |
+
"grad_norm": 0.4120616018772125,
|
| 105 |
+
"learning_rate": 0.0009995548629650466,
|
| 106 |
+
"loss": 1.2282,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 10.238907849829351,
|
| 111 |
+
"grad_norm": 0.5016219019889832,
|
| 112 |
+
"learning_rate": 0.0009993045378942262,
|
| 113 |
+
"loss": 1.2261,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 10.921501706484642,
|
| 118 |
+
"grad_norm": 0.5626952648162842,
|
| 119 |
+
"learning_rate": 0.0009989986480982847,
|
| 120 |
+
"loss": 1.2294,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 11.604095563139932,
|
| 125 |
+
"grad_norm": 0.45088785886764526,
|
| 126 |
+
"learning_rate": 0.000998637231406091,
|
| 127 |
+
"loss": 1.2268,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 12.286689419795222,
|
| 132 |
+
"grad_norm": 0.39686086773872375,
|
| 133 |
+
"learning_rate": 0.0009982203325134304,
|
| 134 |
+
"loss": 1.2254,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 12.969283276450511,
|
| 139 |
+
"grad_norm": 0.3904390037059784,
|
| 140 |
+
"learning_rate": 0.0009977480029774771,
|
| 141 |
+
"loss": 1.2274,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 13.651877133105803,
|
| 146 |
+
"grad_norm": 0.38391488790512085,
|
| 147 |
+
"learning_rate": 0.0009972203012104186,
|
| 148 |
+
"loss": 1.2247,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 14.334470989761092,
|
| 153 |
+
"grad_norm": 0.4345687925815582,
|
| 154 |
+
"learning_rate": 0.0009966372924722312,
|
| 155 |
+
"loss": 1.2252,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 15.017064846416382,
|
| 160 |
+
"grad_norm": 0.34943118691444397,
|
| 161 |
+
"learning_rate": 0.0009959990488626104,
|
| 162 |
+
"loss": 1.2266,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 15.699658703071673,
|
| 167 |
+
"grad_norm": 0.3420363962650299,
|
| 168 |
+
"learning_rate": 0.0009953056493120537,
|
| 169 |
+
"loss": 1.224,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 16.38225255972696,
|
| 174 |
+
"grad_norm": 0.3876005709171295,
|
| 175 |
+
"learning_rate": 0.0009945571795721,
|
| 176 |
+
"loss": 1.2241,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 17.064846416382252,
|
| 181 |
+
"grad_norm": 0.3726520836353302,
|
| 182 |
+
"learning_rate": 0.0009937537322047237,
|
| 183 |
+
"loss": 1.2258,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 17.747440273037544,
|
| 188 |
+
"grad_norm": 0.4071378707885742,
|
| 189 |
+
"learning_rate": 0.0009928954065708895,
|
| 190 |
+
"loss": 1.2235,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 18.43003412969283,
|
| 195 |
+
"grad_norm": 0.443796843290329,
|
| 196 |
+
"learning_rate": 0.0009919823088182628,
|
| 197 |
+
"loss": 1.2346,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 19.112627986348123,
|
| 202 |
+
"grad_norm": 0.3463057279586792,
|
| 203 |
+
"learning_rate": 0.000991014551868084,
|
| 204 |
+
"loss": 1.2292,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 19.795221843003414,
|
| 209 |
+
"grad_norm": 0.31187331676483154,
|
| 210 |
+
"learning_rate": 0.0009899922554012027,
|
| 211 |
+
"loss": 1.2237,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 20.477815699658702,
|
| 216 |
+
"grad_norm": 0.3052064776420593,
|
| 217 |
+
"learning_rate": 0.0009889155458432776,
|
| 218 |
+
"loss": 1.2242,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 21.160409556313994,
|
| 223 |
+
"grad_norm": 0.3126174807548523,
|
| 224 |
+
"learning_rate": 0.000987784556349142,
|
| 225 |
+
"loss": 1.2251,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 21.843003412969285,
|
| 230 |
+
"grad_norm": 0.29788631200790405,
|
| 231 |
+
"learning_rate": 0.0009865994267863351,
|
| 232 |
+
"loss": 1.2224,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 22.525597269624573,
|
| 237 |
+
"grad_norm": 0.2966502606868744,
|
| 238 |
+
"learning_rate": 0.000985360303717807,
|
| 239 |
+
"loss": 1.2233,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 23.208191126279864,
|
| 244 |
+
"grad_norm": 0.31454625725746155,
|
| 245 |
+
"learning_rate": 0.0009840673403837918,
|
| 246 |
+
"loss": 1.2242,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 23.890784982935152,
|
| 251 |
+
"grad_norm": 0.33144181966781616,
|
| 252 |
+
"learning_rate": 0.0009827206966828582,
|
| 253 |
+
"loss": 1.2219,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 24.573378839590443,
|
| 258 |
+
"grad_norm": 0.3603832423686981,
|
| 259 |
+
"learning_rate": 0.000981320539152133,
|
| 260 |
+
"loss": 1.2227,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 25.255972696245735,
|
| 265 |
+
"grad_norm": 0.28107529878616333,
|
| 266 |
+
"learning_rate": 0.0009798670409467075,
|
| 267 |
+
"loss": 1.2236,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 25.938566552901023,
|
| 272 |
+
"grad_norm": 0.3241814076900482,
|
| 273 |
+
"learning_rate": 0.0009783603818182237,
|
| 274 |
+
"loss": 1.2212,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 26.621160409556314,
|
| 279 |
+
"grad_norm": 0.35018429160118103,
|
| 280 |
+
"learning_rate": 0.0009768007480926437,
|
| 281 |
+
"loss": 1.2223,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 27.303754266211605,
|
| 286 |
+
"grad_norm": 0.2918601632118225,
|
| 287 |
+
"learning_rate": 0.0009751883326472076,
|
| 288 |
+
"loss": 1.2228,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 27.986348122866893,
|
| 293 |
+
"grad_norm": 0.29299041628837585,
|
| 294 |
+
"learning_rate": 0.0009735233348865806,
|
| 295 |
+
"loss": 1.2208,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 28.668941979522184,
|
| 300 |
+
"grad_norm": 0.3291133642196655,
|
| 301 |
+
"learning_rate": 0.0009718059607181931,
|
| 302 |
+
"loss": 1.2219,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 29.351535836177476,
|
| 307 |
+
"grad_norm": 0.2924804985523224,
|
| 308 |
+
"learning_rate": 0.0009700364225267769,
|
| 309 |
+
"loss": 1.2222,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 30.034129692832764,
|
| 314 |
+
"grad_norm": 0.3081340193748474,
|
| 315 |
+
"learning_rate": 0.0009682149391480986,
|
| 316 |
+
"loss": 1.2201,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 30.716723549488055,
|
| 321 |
+
"grad_norm": 0.2960987091064453,
|
| 322 |
+
"learning_rate": 0.0009663417358418981,
|
| 323 |
+
"loss": 1.2213,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 31.399317406143346,
|
| 328 |
+
"grad_norm": 0.29421886801719666,
|
| 329 |
+
"learning_rate": 0.0009644170442640303,
|
| 330 |
+
"loss": 1.2201,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 32.081911262798634,
|
| 335 |
+
"grad_norm": 0.3381507098674774,
|
| 336 |
+
"learning_rate": 0.0009624411024378166,
|
| 337 |
+
"loss": 1.2186,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 32.76450511945392,
|
| 342 |
+
"grad_norm": 0.3033551275730133,
|
| 343 |
+
"learning_rate": 0.0009604141547246092,
|
| 344 |
+
"loss": 1.2199,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 33.44709897610922,
|
| 349 |
+
"grad_norm": 0.3156335949897766,
|
| 350 |
+
"learning_rate": 0.0009583364517935713,
|
| 351 |
+
"loss": 1.2189,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 34.129692832764505,
|
| 356 |
+
"grad_norm": 0.33952537178993225,
|
| 357 |
+
"learning_rate": 0.0009562082505906771,
|
| 358 |
+
"loss": 1.2163,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 34.81228668941979,
|
| 363 |
+
"grad_norm": 0.2941412925720215,
|
| 364 |
+
"learning_rate": 0.0009540298143069361,
|
| 365 |
+
"loss": 1.2179,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 35.49488054607509,
|
| 370 |
+
"grad_norm": 0.28668472170829773,
|
| 371 |
+
"learning_rate": 0.0009518014123458442,
|
| 372 |
+
"loss": 1.2158,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 36.177474402730375,
|
| 377 |
+
"grad_norm": 0.2789587080478668,
|
| 378 |
+
"learning_rate": 0.0009495233202900674,
|
| 379 |
+
"loss": 1.213,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 36.86006825938566,
|
| 384 |
+
"grad_norm": 0.3137347102165222,
|
| 385 |
+
"learning_rate": 0.0009471958198673614,
|
| 386 |
+
"loss": 1.214,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 37.54266211604096,
|
| 391 |
+
"grad_norm": 0.25685933232307434,
|
| 392 |
+
"learning_rate": 0.0009448191989157294,
|
| 393 |
+
"loss": 1.21,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 38.225255972696246,
|
| 398 |
+
"grad_norm": 0.3252381980419159,
|
| 399 |
+
"learning_rate": 0.0009423937513478271,
|
| 400 |
+
"loss": 1.2071,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 38.907849829351534,
|
| 405 |
+
"grad_norm": 0.2932053208351135,
|
| 406 |
+
"learning_rate": 0.0009399197771146144,
|
| 407 |
+
"loss": 1.2077,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 39.59044368600683,
|
| 412 |
+
"grad_norm": 0.33668607473373413,
|
| 413 |
+
"learning_rate": 0.0009373975821682607,
|
| 414 |
+
"loss": 1.2019,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 40.27303754266212,
|
| 419 |
+
"grad_norm": 0.31952792406082153,
|
| 420 |
+
"learning_rate": 0.0009348274784243094,
|
| 421 |
+
"loss": 1.1979,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 40.955631399317404,
|
| 426 |
+
"grad_norm": 0.3525752127170563,
|
| 427 |
+
"learning_rate": 0.0009322097837231025,
|
| 428 |
+
"loss": 1.1999,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 41.6382252559727,
|
| 433 |
+
"grad_norm": 0.3389936685562134,
|
| 434 |
+
"learning_rate": 0.0009295448217904745,
|
| 435 |
+
"loss": 1.1923,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 42.32081911262799,
|
| 440 |
+
"grad_norm": 0.3114762008190155,
|
| 441 |
+
"learning_rate": 0.0009268329221977179,
|
| 442 |
+
"loss": 1.19,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 43.003412969283275,
|
| 447 |
+
"grad_norm": 0.3166259229183197,
|
| 448 |
+
"learning_rate": 0.0009240744203208256,
|
| 449 |
+
"loss": 1.1914,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 43.68600682593857,
|
| 454 |
+
"grad_norm": 0.31299668550491333,
|
| 455 |
+
"learning_rate": 0.0009212696572990148,
|
| 456 |
+
"loss": 1.1843,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 44.36860068259386,
|
| 461 |
+
"grad_norm": 0.36728623509407043,
|
| 462 |
+
"learning_rate": 0.0009184189799925403,
|
| 463 |
+
"loss": 1.1828,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 45.051194539249146,
|
| 468 |
+
"grad_norm": 0.25386592745780945,
|
| 469 |
+
"learning_rate": 0.0009155227409397977,
|
| 470 |
+
"loss": 1.1849,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 45.73378839590444,
|
| 475 |
+
"grad_norm": 0.3170667290687561,
|
| 476 |
+
"learning_rate": 0.0009125812983137259,
|
| 477 |
+
"loss": 1.1789,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 46.41638225255973,
|
| 482 |
+
"grad_norm": 0.3354248106479645,
|
| 483 |
+
"learning_rate": 0.0009095950158775128,
|
| 484 |
+
"loss": 1.1781,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 47.098976109215016,
|
| 489 |
+
"grad_norm": 0.2761414647102356,
|
| 490 |
+
"learning_rate": 0.0009065642629396087,
|
| 491 |
+
"loss": 1.1803,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 47.781569965870304,
|
| 496 |
+
"grad_norm": 0.28061142563819885,
|
| 497 |
+
"learning_rate": 0.0009034894143080554,
|
| 498 |
+
"loss": 1.175,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 48.4641638225256,
|
| 503 |
+
"grad_norm": 0.3548315763473511,
|
| 504 |
+
"learning_rate": 0.0009003708502441333,
|
| 505 |
+
"loss": 1.1743,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 49.14675767918089,
|
| 510 |
+
"grad_norm": 0.3077629804611206,
|
| 511 |
+
"learning_rate": 0.0008972089564153364,
|
| 512 |
+
"loss": 1.1762,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 49.829351535836174,
|
| 517 |
+
"grad_norm": 0.3600843846797943,
|
| 518 |
+
"learning_rate": 0.0008940041238476759,
|
| 519 |
+
"loss": 1.1718,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 50.51194539249147,
|
| 524 |
+
"grad_norm": 0.3295997083187103,
|
| 525 |
+
"learning_rate": 0.0008907567488773241,
|
| 526 |
+
"loss": 1.1709,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 51.19453924914676,
|
| 531 |
+
"grad_norm": 0.32083791494369507,
|
| 532 |
+
"learning_rate": 0.0008874672331015996,
|
| 533 |
+
"loss": 1.1727,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 51.877133105802045,
|
| 538 |
+
"grad_norm": 0.32850903272628784,
|
| 539 |
+
"learning_rate": 0.0008841359833293022,
|
| 540 |
+
"loss": 1.1685,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 52.55972696245734,
|
| 545 |
+
"grad_norm": 0.33856746554374695,
|
| 546 |
+
"learning_rate": 0.000880763411530404,
|
| 547 |
+
"loss": 1.168,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 53.24232081911263,
|
| 552 |
+
"grad_norm": 0.34316346049308777,
|
| 553 |
+
"learning_rate": 0.0008773499347851014,
|
| 554 |
+
"loss": 1.1687,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 53.924914675767916,
|
| 559 |
+
"grad_norm": 0.38703614473342896,
|
| 560 |
+
"learning_rate": 0.0008738959752322358,
|
| 561 |
+
"loss": 1.165,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 54.60750853242321,
|
| 566 |
+
"grad_norm": 0.3954058587551117,
|
| 567 |
+
"learning_rate": 0.0008704019600170884,
|
| 568 |
+
"loss": 1.1637,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 55.2901023890785,
|
| 573 |
+
"grad_norm": 0.33417293429374695,
|
| 574 |
+
"learning_rate": 0.0008668683212385556,
|
| 575 |
+
"loss": 1.1635,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 55.972696245733786,
|
| 580 |
+
"grad_norm": 0.3305743932723999,
|
| 581 |
+
"learning_rate": 0.0008632954958957123,
|
| 582 |
+
"loss": 1.1601,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 56.65529010238908,
|
| 587 |
+
"grad_norm": 0.3755279779434204,
|
| 588 |
+
"learning_rate": 0.0008596839258337689,
|
| 589 |
+
"loss": 1.158,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 57.33788395904437,
|
| 594 |
+
"grad_norm": 0.3553084433078766,
|
| 595 |
+
"learning_rate": 0.000856034057689429,
|
| 596 |
+
"loss": 1.1568,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 58.02047781569966,
|
| 601 |
+
"grad_norm": 0.38361021876335144,
|
| 602 |
+
"learning_rate": 0.0008523463428356549,
|
| 603 |
+
"loss": 1.154,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 58.70307167235495,
|
| 608 |
+
"grad_norm": 0.35759079456329346,
|
| 609 |
+
"learning_rate": 0.0008486212373258465,
|
| 610 |
+
"loss": 1.152,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 59.38566552901024,
|
| 615 |
+
"grad_norm": 0.32452645897865295,
|
| 616 |
+
"learning_rate": 0.0008448592018374424,
|
| 617 |
+
"loss": 1.1497,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 60.06825938566553,
|
| 622 |
+
"grad_norm": 0.37652239203453064,
|
| 623 |
+
"learning_rate": 0.0008410607016149491,
|
| 624 |
+
"loss": 1.1461,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 60.75085324232082,
|
| 629 |
+
"grad_norm": 0.35698965191841125,
|
| 630 |
+
"learning_rate": 0.0008372262064124035,
|
| 631 |
+
"loss": 1.1453,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 61.43344709897611,
|
| 636 |
+
"grad_norm": 0.3505449891090393,
|
| 637 |
+
"learning_rate": 0.0008333561904352813,
|
| 638 |
+
"loss": 1.1409,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 62.1160409556314,
|
| 643 |
+
"grad_norm": 0.3646785616874695,
|
| 644 |
+
"learning_rate": 0.000829451132281851,
|
| 645 |
+
"loss": 1.1363,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 62.79863481228669,
|
| 650 |
+
"grad_norm": 0.34945082664489746,
|
| 651 |
+
"learning_rate": 0.0008255115148839877,
|
| 652 |
+
"loss": 1.1371,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 63.48122866894198,
|
| 657 |
+
"grad_norm": 0.34238216280937195,
|
| 658 |
+
"learning_rate": 0.0008215378254474485,
|
| 659 |
+
"loss": 1.1296,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 64.16382252559727,
|
| 664 |
+
"grad_norm": 0.37401944398880005,
|
| 665 |
+
"learning_rate": 0.0008175305553916226,
|
| 666 |
+
"loss": 1.1241,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 64.84641638225256,
|
| 671 |
+
"grad_norm": 0.3763269782066345,
|
| 672 |
+
"learning_rate": 0.000813490200288756,
|
| 673 |
+
"loss": 1.1257,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 65.52901023890784,
|
| 678 |
+
"grad_norm": 0.3553602993488312,
|
| 679 |
+
"learning_rate": 0.0008094172598026662,
|
| 680 |
+
"loss": 1.1147,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 66.21160409556315,
|
| 685 |
+
"grad_norm": 0.36700302362442017,
|
| 686 |
+
"learning_rate": 0.0008053122376269497,
|
| 687 |
+
"loss": 1.1078,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 66.89419795221843,
|
| 692 |
+
"grad_norm": 0.4266979694366455,
|
| 693 |
+
"learning_rate": 0.0008011756414226908,
|
| 694 |
+
"loss": 1.1114,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 67.57679180887372,
|
| 699 |
+
"grad_norm": 0.3603265583515167,
|
| 700 |
+
"learning_rate": 0.0007970079827556795,
|
| 701 |
+
"loss": 1.0967,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 68.25938566552901,
|
| 706 |
+
"grad_norm": 0.39329585433006287,
|
| 707 |
+
"learning_rate": 0.000792809777033148,
|
| 708 |
+
"loss": 1.0898,
|
| 709 |
+
"step": 10000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 68.9419795221843,
|
| 713 |
+
"grad_norm": 0.40609198808670044,
|
| 714 |
+
"learning_rate": 0.0007885815434400306,
|
| 715 |
+
"loss": 1.0944,
|
| 716 |
+
"step": 10100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 69.62457337883959,
|
| 720 |
+
"grad_norm": 0.3622093200683594,
|
| 721 |
+
"learning_rate": 0.0007843238048747572,
|
| 722 |
+
"loss": 1.0761,
|
| 723 |
+
"step": 10200
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 70.30716723549489,
|
| 727 |
+
"grad_norm": 0.39394432306289673,
|
| 728 |
+
"learning_rate": 0.0007800370878845866,
|
| 729 |
+
"loss": 1.0692,
|
| 730 |
+
"step": 10300
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 70.98976109215018,
|
| 734 |
+
"grad_norm": 0.382322758436203,
|
| 735 |
+
"learning_rate": 0.0007757219226004912,
|
| 736 |
+
"loss": 1.0767,
|
| 737 |
+
"step": 10400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 71.67235494880546,
|
| 741 |
+
"grad_norm": 0.3804159462451935,
|
| 742 |
+
"learning_rate": 0.0007713788426715937,
|
| 743 |
+
"loss": 1.0543,
|
| 744 |
+
"step": 10500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 72.35494880546075,
|
| 748 |
+
"grad_norm": 0.3960065245628357,
|
| 749 |
+
"learning_rate": 0.0007670083851991749,
|
| 750 |
+
"loss": 1.0477,
|
| 751 |
+
"step": 10600
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 73.03754266211604,
|
| 755 |
+
"grad_norm": 0.39962372183799744,
|
| 756 |
+
"learning_rate": 0.0007626110906702484,
|
| 757 |
+
"loss": 1.0561,
|
| 758 |
+
"step": 10700
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 73.72013651877133,
|
| 762 |
+
"grad_norm": 0.4019100069999695,
|
| 763 |
+
"learning_rate": 0.0007581875028907215,
|
| 764 |
+
"loss": 1.0329,
|
| 765 |
+
"step": 10800
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 74.40273037542661,
|
| 769 |
+
"grad_norm": 0.42911744117736816,
|
| 770 |
+
"learning_rate": 0.0007537381689181425,
|
| 771 |
+
"loss": 1.0265,
|
| 772 |
+
"step": 10900
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 75.08532423208192,
|
| 776 |
+
"grad_norm": 0.4002211093902588,
|
| 777 |
+
"learning_rate": 0.0007492636389940478,
|
| 778 |
+
"loss": 1.0332,
|
| 779 |
+
"step": 11000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 75.7679180887372,
|
| 783 |
+
"grad_norm": 0.40011659264564514,
|
| 784 |
+
"learning_rate": 0.0007447644664759138,
|
| 785 |
+
"loss": 1.0127,
|
| 786 |
+
"step": 11100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 76.45051194539249,
|
| 790 |
+
"grad_norm": 0.43485236167907715,
|
| 791 |
+
"learning_rate": 0.0007402412077687246,
|
| 792 |
+
"loss": 1.0047,
|
| 793 |
+
"step": 11200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 77.13310580204778,
|
| 797 |
+
"grad_norm": 0.4095554053783417,
|
| 798 |
+
"learning_rate": 0.0007356944222561622,
|
| 799 |
+
"loss": 1.0114,
|
| 800 |
+
"step": 11300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 77.81569965870307,
|
| 804 |
+
"grad_norm": 0.39631539583206177,
|
| 805 |
+
"learning_rate": 0.0007311246722314281,
|
| 806 |
+
"loss": 0.9916,
|
| 807 |
+
"step": 11400
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 78.49829351535836,
|
| 811 |
+
"grad_norm": 0.4598711133003235,
|
| 812 |
+
"learning_rate": 0.0007265325228277065,
|
| 813 |
+
"loss": 0.9822,
|
| 814 |
+
"step": 11500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 79.18088737201366,
|
| 818 |
+
"grad_norm": 0.39505425095558167,
|
| 819 |
+
"learning_rate": 0.0007219185419482739,
|
| 820 |
+
"loss": 0.9883,
|
| 821 |
+
"step": 11600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 79.86348122866895,
|
| 825 |
+
"grad_norm": 0.3844675123691559,
|
| 826 |
+
"learning_rate": 0.0007172833001962683,
|
| 827 |
+
"loss": 0.9701,
|
| 828 |
+
"step": 11700
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 80.54607508532423,
|
| 832 |
+
"grad_norm": 0.48971033096313477,
|
| 833 |
+
"learning_rate": 0.000712627370804124,
|
| 834 |
+
"loss": 0.9597,
|
| 835 |
+
"step": 11800
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 81.22866894197952,
|
| 839 |
+
"grad_norm": 0.40114977955818176,
|
| 840 |
+
"learning_rate": 0.0007079513295626796,
|
| 841 |
+
"loss": 0.9643,
|
| 842 |
+
"step": 11900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 81.91126279863481,
|
| 846 |
+
"grad_norm": 0.42664656043052673,
|
| 847 |
+
"learning_rate": 0.0007032557547499727,
|
| 848 |
+
"loss": 0.9482,
|
| 849 |
+
"step": 12000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 82.5938566552901,
|
| 853 |
+
"grad_norm": 0.5133378505706787,
|
| 854 |
+
"learning_rate": 0.0006985412270597231,
|
| 855 |
+
"loss": 0.9363,
|
| 856 |
+
"step": 12100
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 83.2764505119454,
|
| 860 |
+
"grad_norm": 0.43057283759117126,
|
| 861 |
+
"learning_rate": 0.000693808329529522,
|
| 862 |
+
"loss": 0.9409,
|
| 863 |
+
"step": 12200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 83.95904436860069,
|
| 867 |
+
"grad_norm": 0.4190930426120758,
|
| 868 |
+
"learning_rate": 0.0006890576474687264,
|
| 869 |
+
"loss": 0.9255,
|
| 870 |
+
"step": 12300
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 84.64163822525597,
|
| 874 |
+
"grad_norm": 0.43727907538414,
|
| 875 |
+
"learning_rate": 0.000684289768386076,
|
| 876 |
+
"loss": 0.9119,
|
| 877 |
+
"step": 12400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 85.32423208191126,
|
| 881 |
+
"grad_norm": 0.4994739890098572,
|
| 882 |
+
"learning_rate": 0.000679505281917037,
|
| 883 |
+
"loss": 0.914,
|
| 884 |
+
"step": 12500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 86.00682593856655,
|
| 888 |
+
"grad_norm": 0.45477139949798584,
|
| 889 |
+
"learning_rate": 0.0006747047797508828,
|
| 890 |
+
"loss": 0.9017,
|
| 891 |
+
"step": 12600
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 86.68941979522184,
|
| 895 |
+
"grad_norm": 0.4953708052635193,
|
| 896 |
+
"learning_rate": 0.0006698888555575204,
|
| 897 |
+
"loss": 0.8881,
|
| 898 |
+
"step": 12700
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 87.37201365187714,
|
| 902 |
+
"grad_norm": 0.4449358284473419,
|
| 903 |
+
"learning_rate": 0.000665058104914073,
|
| 904 |
+
"loss": 0.8875,
|
| 905 |
+
"step": 12800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 88.05460750853243,
|
| 909 |
+
"grad_norm": 0.4819263815879822,
|
| 910 |
+
"learning_rate": 0.0006602131252312249,
|
| 911 |
+
"loss": 0.8748,
|
| 912 |
+
"step": 12900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 88.73720136518772,
|
| 916 |
+
"grad_norm": 0.5412826538085938,
|
| 917 |
+
"learning_rate": 0.0006553545156793422,
|
| 918 |
+
"loss": 0.8655,
|
| 919 |
+
"step": 13000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 89.419795221843,
|
| 923 |
+
"grad_norm": 0.4753580391407013,
|
| 924 |
+
"learning_rate": 0.0006504828771143733,
|
| 925 |
+
"loss": 0.8617,
|
| 926 |
+
"step": 13100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 90.10238907849829,
|
| 930 |
+
"grad_norm": 0.545566737651825,
|
| 931 |
+
"learning_rate": 0.0006455988120035418,
|
| 932 |
+
"loss": 0.85,
|
| 933 |
+
"step": 13200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 90.78498293515358,
|
| 937 |
+
"grad_norm": 0.5255779027938843,
|
| 938 |
+
"learning_rate": 0.0006407029243508423,
|
| 939 |
+
"loss": 0.8449,
|
| 940 |
+
"step": 13300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 91.46757679180887,
|
| 944 |
+
"grad_norm": 0.43862029910087585,
|
| 945 |
+
"learning_rate": 0.0006357958196223419,
|
| 946 |
+
"loss": 0.8371,
|
| 947 |
+
"step": 13400
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 92.15017064846417,
|
| 951 |
+
"grad_norm": 0.5360426902770996,
|
| 952 |
+
"learning_rate": 0.0006308781046713046,
|
| 953 |
+
"loss": 0.8257,
|
| 954 |
+
"step": 13500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 92.83276450511946,
|
| 958 |
+
"grad_norm": 0.5451890230178833,
|
| 959 |
+
"learning_rate": 0.0006259503876631428,
|
| 960 |
+
"loss": 0.8247,
|
| 961 |
+
"step": 13600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 93.51535836177474,
|
| 965 |
+
"grad_norm": 0.4524667263031006,
|
| 966 |
+
"learning_rate": 0.0006210132780002057,
|
| 967 |
+
"loss": 0.8147,
|
| 968 |
+
"step": 13700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 94.19795221843003,
|
| 972 |
+
"grad_norm": 0.6203665733337402,
|
| 973 |
+
"learning_rate": 0.0006160673862464167,
|
| 974 |
+
"loss": 0.8022,
|
| 975 |
+
"step": 13800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 94.88054607508532,
|
| 979 |
+
"grad_norm": 0.5576283931732178,
|
| 980 |
+
"learning_rate": 0.0006111133240517648,
|
| 981 |
+
"loss": 0.8057,
|
| 982 |
+
"step": 13900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 95.56313993174061,
|
| 986 |
+
"grad_norm": 0.48380613327026367,
|
| 987 |
+
"learning_rate": 0.000606151704076664,
|
| 988 |
+
"loss": 0.7906,
|
| 989 |
+
"step": 14000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 96.24573378839591,
|
| 993 |
+
"grad_norm": 0.5328605771064758,
|
| 994 |
+
"learning_rate": 0.0006011831399161856,
|
| 995 |
+
"loss": 0.7823,
|
| 996 |
+
"step": 14100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 96.9283276450512,
|
| 1000 |
+
"grad_norm": 0.5395786762237549,
|
| 1001 |
+
"learning_rate": 0.0005962082460241767,
|
| 1002 |
+
"loss": 0.7883,
|
| 1003 |
+
"step": 14200
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 97.61092150170649,
|
| 1007 |
+
"grad_norm": 0.48329558968544006,
|
| 1008 |
+
"learning_rate": 0.0005912276376372715,
|
| 1009 |
+
"loss": 0.7707,
|
| 1010 |
+
"step": 14300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 98.29351535836177,
|
| 1014 |
+
"grad_norm": 0.5496215224266052,
|
| 1015 |
+
"learning_rate": 0.0005862419306988061,
|
| 1016 |
+
"loss": 0.7636,
|
| 1017 |
+
"step": 14400
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 98.97610921501706,
|
| 1021 |
+
"grad_norm": 0.6007915139198303,
|
| 1022 |
+
"learning_rate": 0.0005812517417826456,
|
| 1023 |
+
"loss": 0.7712,
|
| 1024 |
+
"step": 14500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 99.65870307167235,
|
| 1028 |
+
"grad_norm": 0.4981062710285187,
|
| 1029 |
+
"learning_rate": 0.000576257688016934,
|
| 1030 |
+
"loss": 0.7529,
|
| 1031 |
+
"step": 14600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 100.34129692832765,
|
| 1035 |
+
"grad_norm": 0.5475151538848877,
|
| 1036 |
+
"learning_rate": 0.0005712603870077736,
|
| 1037 |
+
"loss": 0.7468,
|
| 1038 |
+
"step": 14700
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 101.02389078498294,
|
| 1042 |
+
"grad_norm": 0.4879900813102722,
|
| 1043 |
+
"learning_rate": 0.0005662604567628494,
|
| 1044 |
+
"loss": 0.7568,
|
| 1045 |
+
"step": 14800
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 101.70648464163823,
|
| 1049 |
+
"grad_norm": 0.4963570535182953,
|
| 1050 |
+
"learning_rate": 0.0005612585156149982,
|
| 1051 |
+
"loss": 0.7386,
|
| 1052 |
+
"step": 14900
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 102.38907849829351,
|
| 1056 |
+
"grad_norm": 0.5414558053016663,
|
| 1057 |
+
"learning_rate": 0.0005562551821457418,
|
| 1058 |
+
"loss": 0.7326,
|
| 1059 |
+
"step": 15000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 103.0716723549488,
|
| 1063 |
+
"grad_norm": 0.47427278757095337,
|
| 1064 |
+
"learning_rate": 0.0005512510751087889,
|
| 1065 |
+
"loss": 0.7415,
|
| 1066 |
+
"step": 15100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 103.75426621160409,
|
| 1070 |
+
"grad_norm": 0.48589903116226196,
|
| 1071 |
+
"learning_rate": 0.0005462468133535131,
|
| 1072 |
+
"loss": 0.7267,
|
| 1073 |
+
"step": 15200
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 104.43686006825939,
|
| 1077 |
+
"grad_norm": 0.5541400909423828,
|
| 1078 |
+
"learning_rate": 0.0005412430157484224,
|
| 1079 |
+
"loss": 0.7206,
|
| 1080 |
+
"step": 15300
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 105.11945392491468,
|
| 1084 |
+
"grad_norm": 0.4941166341304779,
|
| 1085 |
+
"learning_rate": 0.0005362403011046234,
|
| 1086 |
+
"loss": 0.7289,
|
| 1087 |
+
"step": 15400
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 105.80204778156997,
|
| 1091 |
+
"grad_norm": 0.4855073094367981,
|
| 1092 |
+
"learning_rate": 0.0005312392880992954,
|
| 1093 |
+
"loss": 0.716,
|
| 1094 |
+
"step": 15500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 106.48464163822526,
|
| 1098 |
+
"grad_norm": 0.5357630252838135,
|
| 1099 |
+
"learning_rate": 0.0005262405951991788,
|
| 1100 |
+
"loss": 0.7111,
|
| 1101 |
+
"step": 15600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 107.16723549488054,
|
| 1105 |
+
"grad_norm": 0.5147479772567749,
|
| 1106 |
+
"learning_rate": 0.0005212448405840905,
|
| 1107 |
+
"loss": 0.7181,
|
| 1108 |
+
"step": 15700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 107.84982935153583,
|
| 1112 |
+
"grad_norm": 0.47526028752326965,
|
| 1113 |
+
"learning_rate": 0.0005162526420704744,
|
| 1114 |
+
"loss": 0.7072,
|
| 1115 |
+
"step": 15800
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 108.53242320819113,
|
| 1119 |
+
"grad_norm": 0.528582751750946,
|
| 1120 |
+
"learning_rate": 0.0005112646170349981,
|
| 1121 |
+
"loss": 0.7021,
|
| 1122 |
+
"step": 15900
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 109.21501706484642,
|
| 1126 |
+
"grad_norm": 0.4804699420928955,
|
| 1127 |
+
"learning_rate": 0.0005062813823382014,
|
| 1128 |
+
"loss": 0.7093,
|
| 1129 |
+
"step": 16000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 109.89761092150171,
|
| 1133 |
+
"grad_norm": 0.5126569271087646,
|
| 1134 |
+
"learning_rate": 0.000501303554248212,
|
| 1135 |
+
"loss": 0.7008,
|
| 1136 |
+
"step": 16100
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 110.580204778157,
|
| 1140 |
+
"grad_norm": 0.5352160930633545,
|
| 1141 |
+
"learning_rate": 0.0004963317483645311,
|
| 1142 |
+
"loss": 0.6946,
|
| 1143 |
+
"step": 16200
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 111.26279863481228,
|
| 1147 |
+
"grad_norm": 0.5465342402458191,
|
| 1148 |
+
"learning_rate": 0.0004913665795419041,
|
| 1149 |
+
"loss": 0.7008,
|
| 1150 |
+
"step": 16300
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 111.94539249146757,
|
| 1154 |
+
"grad_norm": 0.45367056131362915,
|
| 1155 |
+
"learning_rate": 0.00048640866181428265,
|
| 1156 |
+
"loss": 0.6939,
|
| 1157 |
+
"step": 16400
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 112.62798634812286,
|
| 1161 |
+
"grad_norm": 0.6308451294898987,
|
| 1162 |
+
"learning_rate": 0.00048145860831888776,
|
| 1163 |
+
"loss": 0.6885,
|
| 1164 |
+
"step": 16500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 113.31058020477816,
|
| 1168 |
+
"grad_norm": 0.4779800772666931,
|
| 1169 |
+
"learning_rate": 0.00047651703122038395,
|
| 1170 |
+
"loss": 0.6939,
|
| 1171 |
+
"step": 16600
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 113.99317406143345,
|
| 1175 |
+
"grad_norm": 0.4980040192604065,
|
| 1176 |
+
"learning_rate": 0.0004715845416351744,
|
| 1177 |
+
"loss": 0.6879,
|
| 1178 |
+
"step": 16700
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 114.67576791808874,
|
| 1182 |
+
"grad_norm": 0.5260293483734131,
|
| 1183 |
+
"learning_rate": 0.00046666174955582507,
|
| 1184 |
+
"loss": 0.6835,
|
| 1185 |
+
"step": 16800
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 115.35836177474403,
|
| 1189 |
+
"grad_norm": 0.47831013798713684,
|
| 1190 |
+
"learning_rate": 0.0004617492637756282,
|
| 1191 |
+
"loss": 0.6877,
|
| 1192 |
+
"step": 16900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 116.04095563139931,
|
| 1196 |
+
"grad_norm": 0.5765718221664429,
|
| 1197 |
+
"learning_rate": 0.0004568476918133129,
|
| 1198 |
+
"loss": 0.6813,
|
| 1199 |
+
"step": 17000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 116.7235494880546,
|
| 1203 |
+
"grad_norm": 0.5322794318199158,
|
| 1204 |
+
"learning_rate": 0.00045195763983791573,
|
| 1205 |
+
"loss": 0.679,
|
| 1206 |
+
"step": 17100
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 117.4061433447099,
|
| 1210 |
+
"grad_norm": 0.514970064163208,
|
| 1211 |
+
"learning_rate": 0.00044707971259381584,
|
| 1212 |
+
"loss": 0.68,
|
| 1213 |
+
"step": 17200
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 118.08873720136519,
|
| 1217 |
+
"grad_norm": 0.539171576499939,
|
| 1218 |
+
"learning_rate": 0.00044221451332594785,
|
| 1219 |
+
"loss": 0.6769,
|
| 1220 |
+
"step": 17300
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 118.77133105802048,
|
| 1224 |
+
"grad_norm": 0.5568128824234009,
|
| 1225 |
+
"learning_rate": 0.00043736264370519863,
|
| 1226 |
+
"loss": 0.6757,
|
| 1227 |
+
"step": 17400
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 119.45392491467577,
|
| 1231 |
+
"grad_norm": 0.4861965477466583,
|
| 1232 |
+
"learning_rate": 0.0004325247037540011,
|
| 1233 |
+
"loss": 0.6749,
|
| 1234 |
+
"step": 17500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 120.13651877133105,
|
| 1238 |
+
"grad_norm": 0.49631592631340027,
|
| 1239 |
+
"learning_rate": 0.00042770129177212914,
|
| 1240 |
+
"loss": 0.671,
|
| 1241 |
+
"step": 17600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 120.81911262798634,
|
| 1245 |
+
"grad_norm": 0.5209415555000305,
|
| 1246 |
+
"learning_rate": 0.00042289300426270806,
|
| 1247 |
+
"loss": 0.6714,
|
| 1248 |
+
"step": 17700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 121.50170648464164,
|
| 1252 |
+
"grad_norm": 0.4455106556415558,
|
| 1253 |
+
"learning_rate": 0.0004181004358584447,
|
| 1254 |
+
"loss": 0.6699,
|
| 1255 |
+
"step": 17800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 122.18430034129693,
|
| 1259 |
+
"grad_norm": 0.46950340270996094,
|
| 1260 |
+
"learning_rate": 0.00041332417924809175,
|
| 1261 |
+
"loss": 0.6654,
|
| 1262 |
+
"step": 17900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 122.86689419795222,
|
| 1266 |
+
"grad_norm": 0.5346481800079346,
|
| 1267 |
+
"learning_rate": 0.00040856482510314985,
|
| 1268 |
+
"loss": 0.6681,
|
| 1269 |
+
"step": 18000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 123.54948805460751,
|
| 1273 |
+
"grad_norm": 0.47252893447875977,
|
| 1274 |
+
"learning_rate": 0.0004038229620048207,
|
| 1275 |
+
"loss": 0.6648,
|
| 1276 |
+
"step": 18100
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 124.2320819112628,
|
| 1280 |
+
"grad_norm": 0.5372588634490967,
|
| 1281 |
+
"learning_rate": 0.00039909917637121796,
|
| 1282 |
+
"loss": 0.6611,
|
| 1283 |
+
"step": 18200
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 124.91467576791808,
|
| 1287 |
+
"grad_norm": 0.5461378693580627,
|
| 1288 |
+
"learning_rate": 0.00039439405238484614,
|
| 1289 |
+
"loss": 0.6637,
|
| 1290 |
+
"step": 18300
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 125.59726962457339,
|
| 1294 |
+
"grad_norm": 0.4744736850261688,
|
| 1295 |
+
"learning_rate": 0.0003897081719203558,
|
| 1296 |
+
"loss": 0.659,
|
| 1297 |
+
"step": 18400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 126.27986348122867,
|
| 1301 |
+
"grad_norm": 0.5168038606643677,
|
| 1302 |
+
"learning_rate": 0.0003850421144725844,
|
| 1303 |
+
"loss": 0.6559,
|
| 1304 |
+
"step": 18500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 126.96245733788396,
|
| 1308 |
+
"grad_norm": 0.5486589670181274,
|
| 1309 |
+
"learning_rate": 0.00038039645708488955,
|
| 1310 |
+
"loss": 0.6606,
|
| 1311 |
+
"step": 18600
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 127.64505119453925,
|
| 1315 |
+
"grad_norm": 0.44690069556236267,
|
| 1316 |
+
"learning_rate": 0.0003757717742777895,
|
| 1317 |
+
"loss": 0.6538,
|
| 1318 |
+
"step": 18700
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 128.32764505119454,
|
| 1322 |
+
"grad_norm": 0.5112203359603882,
|
| 1323 |
+
"learning_rate": 0.00037116863797791145,
|
| 1324 |
+
"loss": 0.6523,
|
| 1325 |
+
"step": 18800
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 129.01023890784984,
|
| 1329 |
+
"grad_norm": 0.479837030172348,
|
| 1330 |
+
"learning_rate": 0.0003665876174472628,
|
| 1331 |
+
"loss": 0.6563,
|
| 1332 |
+
"step": 18900
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 129.6928327645051,
|
| 1336 |
+
"grad_norm": 0.4698416292667389,
|
| 1337 |
+
"learning_rate": 0.00036202927921283144,
|
| 1338 |
+
"loss": 0.6495,
|
| 1339 |
+
"step": 19000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 130.37542662116041,
|
| 1343 |
+
"grad_norm": 0.5283064246177673,
|
| 1344 |
+
"learning_rate": 0.0003574941869965248,
|
| 1345 |
+
"loss": 0.6474,
|
| 1346 |
+
"step": 19100
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 131.0580204778157,
|
| 1350 |
+
"grad_norm": 0.4855287969112396,
|
| 1351 |
+
"learning_rate": 0.00035298290164545443,
|
| 1352 |
+
"loss": 0.652,
|
| 1353 |
+
"step": 19200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 131.740614334471,
|
| 1357 |
+
"grad_norm": 0.4482058882713318,
|
| 1358 |
+
"learning_rate": 0.00034849598106257795,
|
| 1359 |
+
"loss": 0.6454,
|
| 1360 |
+
"step": 19300
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 132.4232081911263,
|
| 1364 |
+
"grad_norm": 0.487212210893631,
|
| 1365 |
+
"learning_rate": 0.00034403398013770347,
|
| 1366 |
+
"loss": 0.643,
|
| 1367 |
+
"step": 19400
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 133.10580204778157,
|
| 1371 |
+
"grad_norm": 0.4655930697917938,
|
| 1372 |
+
"learning_rate": 0.0003395974506788674,
|
| 1373 |
+
"loss": 0.6477,
|
| 1374 |
+
"step": 19500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 133.78839590443687,
|
| 1378 |
+
"grad_norm": 0.4315454661846161,
|
| 1379 |
+
"learning_rate": 0.00033518694134409367,
|
| 1380 |
+
"loss": 0.6411,
|
| 1381 |
+
"step": 19600
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 134.47098976109214,
|
| 1385 |
+
"grad_norm": 0.505107581615448,
|
| 1386 |
+
"learning_rate": 0.00033080299757354216,
|
| 1387 |
+
"loss": 0.6396,
|
| 1388 |
+
"step": 19700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 135.15358361774744,
|
| 1392 |
+
"grad_norm": 0.47746554017066956,
|
| 1393 |
+
"learning_rate": 0.00032644616152205385,
|
| 1394 |
+
"loss": 0.6436,
|
| 1395 |
+
"step": 19800
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 135.83617747440272,
|
| 1399 |
+
"grad_norm": 0.44861286878585815,
|
| 1400 |
+
"learning_rate": 0.0003221169719921052,
|
| 1401 |
+
"loss": 0.6382,
|
| 1402 |
+
"step": 19900
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 136.51877133105802,
|
| 1406 |
+
"grad_norm": 0.4924197196960449,
|
| 1407 |
+
"learning_rate": 0.00031781596436717465,
|
| 1408 |
+
"loss": 0.636,
|
| 1409 |
+
"step": 20000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 137.20136518771332,
|
| 1413 |
+
"grad_norm": 0.4533005952835083,
|
| 1414 |
+
"learning_rate": 0.0003135436705455319,
|
| 1415 |
+
"loss": 0.6388,
|
| 1416 |
+
"step": 20100
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 137.8839590443686,
|
| 1420 |
+
"grad_norm": 0.4372631311416626,
|
| 1421 |
+
"learning_rate": 0.0003093006188744598,
|
| 1422 |
+
"loss": 0.6349,
|
| 1423 |
+
"step": 20200
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 138.5665529010239,
|
| 1427 |
+
"grad_norm": 0.5831769108772278,
|
| 1428 |
+
"learning_rate": 0.0003050873340849148,
|
| 1429 |
+
"loss": 0.6323,
|
| 1430 |
+
"step": 20300
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 139.24914675767917,
|
| 1434 |
+
"grad_norm": 0.41959720849990845,
|
| 1435 |
+
"learning_rate": 0.00030090433722663367,
|
| 1436 |
+
"loss": 0.6357,
|
| 1437 |
+
"step": 20400
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 139.93174061433447,
|
| 1441 |
+
"grad_norm": 0.47732993960380554,
|
| 1442 |
+
"learning_rate": 0.00029675214560369637,
|
| 1443 |
+
"loss": 0.6323,
|
| 1444 |
+
"step": 20500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 140.61433447098977,
|
| 1448 |
+
"grad_norm": 0.4794413447380066,
|
| 1449 |
+
"learning_rate": 0.0002926312727105518,
|
| 1450 |
+
"loss": 0.6303,
|
| 1451 |
+
"step": 20600
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 141.29692832764505,
|
| 1455 |
+
"grad_norm": 0.42916139960289,
|
| 1456 |
+
"learning_rate": 0.00028854222816851547,
|
| 1457 |
+
"loss": 0.6327,
|
| 1458 |
+
"step": 20700
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 141.97952218430035,
|
| 1462 |
+
"grad_norm": 0.42981767654418945,
|
| 1463 |
+
"learning_rate": 0.00028448551766274466,
|
| 1464 |
+
"loss": 0.6285,
|
| 1465 |
+
"step": 20800
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 142.66211604095562,
|
| 1469 |
+
"grad_norm": 0.48353445529937744,
|
| 1470 |
+
"learning_rate": 0.0002804616428797021,
|
| 1471 |
+
"loss": 0.6272,
|
| 1472 |
+
"step": 20900
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 143.34470989761093,
|
| 1476 |
+
"grad_norm": 0.4736007750034332,
|
| 1477 |
+
"learning_rate": 0.0002764711014451121,
|
| 1478 |
+
"loss": 0.6288,
|
| 1479 |
+
"step": 21000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 144.0273037542662,
|
| 1483 |
+
"grad_norm": 0.47045692801475525,
|
| 1484 |
+
"learning_rate": 0.0002725143868624209,
|
| 1485 |
+
"loss": 0.6257,
|
| 1486 |
+
"step": 21100
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 144.7098976109215,
|
| 1490 |
+
"grad_norm": 0.4963953495025635,
|
| 1491 |
+
"learning_rate": 0.0002685919884517659,
|
| 1492 |
+
"loss": 0.6251,
|
| 1493 |
+
"step": 21200
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 145.3924914675768,
|
| 1497 |
+
"grad_norm": 0.4388594925403595,
|
| 1498 |
+
"learning_rate": 0.00026470439128946214,
|
| 1499 |
+
"loss": 0.6249,
|
| 1500 |
+
"step": 21300
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 146.07508532423208,
|
| 1504 |
+
"grad_norm": 0.4757547378540039,
|
| 1505 |
+
"learning_rate": 0.0002608520761480135,
|
| 1506 |
+
"loss": 0.6241,
|
| 1507 |
+
"step": 21400
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 146.75767918088738,
|
| 1511 |
+
"grad_norm": 0.46713003516197205,
|
| 1512 |
+
"learning_rate": 0.00025703551943665636,
|
| 1513 |
+
"loss": 0.6229,
|
| 1514 |
+
"step": 21500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 147.44027303754265,
|
| 1518 |
+
"grad_norm": 0.43607574701309204,
|
| 1519 |
+
"learning_rate": 0.00025325519314244374,
|
| 1520 |
+
"loss": 0.623,
|
| 1521 |
+
"step": 21600
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 148.12286689419795,
|
| 1525 |
+
"grad_norm": 0.4766788184642792,
|
| 1526 |
+
"learning_rate": 0.00024951156477187413,
|
| 1527 |
+
"loss": 0.6212,
|
| 1528 |
+
"step": 21700
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 148.80546075085323,
|
| 1532 |
+
"grad_norm": 0.49822133779525757,
|
| 1533 |
+
"learning_rate": 0.00024580509729307673,
|
| 1534 |
+
"loss": 0.6215,
|
| 1535 |
+
"step": 21800
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 149.48805460750853,
|
| 1539 |
+
"grad_norm": 0.4069366455078125,
|
| 1540 |
+
"learning_rate": 0.0002421362490785569,
|
| 1541 |
+
"loss": 0.6199,
|
| 1542 |
+
"step": 21900
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 150.17064846416383,
|
| 1546 |
+
"grad_norm": 0.46552979946136475,
|
| 1547 |
+
"learning_rate": 0.00023850547384850974,
|
| 1548 |
+
"loss": 0.6194,
|
| 1549 |
+
"step": 22000
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 150.8532423208191,
|
| 1553 |
+
"grad_norm": 0.45079755783081055,
|
| 1554 |
+
"learning_rate": 0.00023491322061470915,
|
| 1555 |
+
"loss": 0.6202,
|
| 1556 |
+
"step": 22100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 151.5358361774744,
|
| 1560 |
+
"grad_norm": 0.46759602427482605,
|
| 1561 |
+
"learning_rate": 0.00023135993362497948,
|
| 1562 |
+
"loss": 0.6181,
|
| 1563 |
+
"step": 22200
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 152.21843003412968,
|
| 1567 |
+
"grad_norm": 0.45771974325180054,
|
| 1568 |
+
"learning_rate": 0.00022784605230825666,
|
| 1569 |
+
"loss": 0.6164,
|
| 1570 |
+
"step": 22300
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 152.90102389078498,
|
| 1574 |
+
"grad_norm": 0.5069631338119507,
|
| 1575 |
+
"learning_rate": 0.00022437201122024375,
|
| 1576 |
+
"loss": 0.6187,
|
| 1577 |
+
"step": 22400
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 153.5836177474403,
|
| 1581 |
+
"grad_norm": 0.43779122829437256,
|
| 1582 |
+
"learning_rate": 0.00022093823998967115,
|
| 1583 |
+
"loss": 0.6164,
|
| 1584 |
+
"step": 22500
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 154.26621160409556,
|
| 1588 |
+
"grad_norm": 0.477224200963974,
|
| 1589 |
+
"learning_rate": 0.00021754516326516458,
|
| 1590 |
+
"loss": 0.6151,
|
| 1591 |
+
"step": 22600
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 154.94880546075086,
|
| 1595 |
+
"grad_norm": 0.47925087809562683,
|
| 1596 |
+
"learning_rate": 0.00021419320066272973,
|
| 1597 |
+
"loss": 0.6166,
|
| 1598 |
+
"step": 22700
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 155.63139931740614,
|
| 1602 |
+
"grad_norm": 0.4455689787864685,
|
| 1603 |
+
"learning_rate": 0.00021088276671385882,
|
| 1604 |
+
"loss": 0.6139,
|
| 1605 |
+
"step": 22800
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 156.31399317406144,
|
| 1609 |
+
"grad_norm": 0.4499022364616394,
|
| 1610 |
+
"learning_rate": 0.00020761427081426677,
|
| 1611 |
+
"loss": 0.6132,
|
| 1612 |
+
"step": 22900
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 156.9965870307167,
|
| 1616 |
+
"grad_norm": 0.4519612193107605,
|
| 1617 |
+
"learning_rate": 0.00020438811717326112,
|
| 1618 |
+
"loss": 0.6154,
|
| 1619 |
+
"step": 23000
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 157.679180887372,
|
| 1623 |
+
"grad_norm": 0.4130764603614807,
|
| 1624 |
+
"learning_rate": 0.0002012047047637547,
|
| 1625 |
+
"loss": 0.6122,
|
| 1626 |
+
"step": 23100
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 158.36177474402731,
|
| 1630 |
+
"grad_norm": 0.44488415122032166,
|
| 1631 |
+
"learning_rate": 0.0001980644272729254,
|
| 1632 |
+
"loss": 0.6132,
|
| 1633 |
+
"step": 23200
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 159.0443686006826,
|
| 1637 |
+
"grad_norm": 0.47042757272720337,
|
| 1638 |
+
"learning_rate": 0.00019496767305352946,
|
| 1639 |
+
"loss": 0.6135,
|
| 1640 |
+
"step": 23300
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 159.7269624573379,
|
| 1644 |
+
"grad_norm": 0.3974725604057312,
|
| 1645 |
+
"learning_rate": 0.00019191482507587376,
|
| 1646 |
+
"loss": 0.6107,
|
| 1647 |
+
"step": 23400
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 160.40955631399316,
|
| 1651 |
+
"grad_norm": 0.46765318512916565,
|
| 1652 |
+
"learning_rate": 0.00018890626088045593,
|
| 1653 |
+
"loss": 0.6109,
|
| 1654 |
+
"step": 23500
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 161.09215017064847,
|
| 1658 |
+
"grad_norm": 0.42309921979904175,
|
| 1659 |
+
"learning_rate": 0.00018594235253127373,
|
| 1660 |
+
"loss": 0.6122,
|
| 1661 |
+
"step": 23600
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 161.77474402730377,
|
| 1665 |
+
"grad_norm": 0.3989265561103821,
|
| 1666 |
+
"learning_rate": 0.00018302346656981234,
|
| 1667 |
+
"loss": 0.6099,
|
| 1668 |
+
"step": 23700
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 162.45733788395904,
|
| 1672 |
+
"grad_norm": 0.4663792550563812,
|
| 1673 |
+
"learning_rate": 0.0001801499639697153,
|
| 1674 |
+
"loss": 0.6103,
|
| 1675 |
+
"step": 23800
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 163.13993174061434,
|
| 1679 |
+
"grad_norm": 0.3913436233997345,
|
| 1680 |
+
"learning_rate": 0.0001773222000921435,
|
| 1681 |
+
"loss": 0.6106,
|
| 1682 |
+
"step": 23900
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 163.82252559726962,
|
| 1686 |
+
"grad_norm": 0.41263365745544434,
|
| 1687 |
+
"learning_rate": 0.00017454052464182778,
|
| 1688 |
+
"loss": 0.6092,
|
| 1689 |
+
"step": 24000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 164.50511945392492,
|
| 1693 |
+
"grad_norm": 0.4633448123931885,
|
| 1694 |
+
"learning_rate": 0.0001718052816238222,
|
| 1695 |
+
"loss": 0.608,
|
| 1696 |
+
"step": 24100
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 165.1877133105802,
|
| 1700 |
+
"grad_norm": 0.3786383271217346,
|
| 1701 |
+
"learning_rate": 0.00016911680930096085,
|
| 1702 |
+
"loss": 0.6094,
|
| 1703 |
+
"step": 24200
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 165.8703071672355,
|
| 1707 |
+
"grad_norm": 0.4065646529197693,
|
| 1708 |
+
"learning_rate": 0.00016647544015202623,
|
| 1709 |
+
"loss": 0.608,
|
| 1710 |
+
"step": 24300
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 166.5529010238908,
|
| 1714 |
+
"grad_norm": 0.43729546666145325,
|
| 1715 |
+
"learning_rate": 0.00016388150083063117,
|
| 1716 |
+
"loss": 0.6072,
|
| 1717 |
+
"step": 24400
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 167.23549488054607,
|
| 1721 |
+
"grad_norm": 0.4111308157444,
|
| 1722 |
+
"learning_rate": 0.00016133531212482325,
|
| 1723 |
+
"loss": 0.6078,
|
| 1724 |
+
"step": 24500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 167.91808873720137,
|
| 1728 |
+
"grad_norm": 0.39553767442703247,
|
| 1729 |
+
"learning_rate": 0.00015883718891741258,
|
| 1730 |
+
"loss": 0.6066,
|
| 1731 |
+
"step": 24600
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 168.60068259385665,
|
| 1735 |
+
"grad_norm": 0.4074500501155853,
|
| 1736 |
+
"learning_rate": 0.00015638744014703128,
|
| 1737 |
+
"loss": 0.6068,
|
| 1738 |
+
"step": 24700
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 169.28327645051195,
|
| 1742 |
+
"grad_norm": 0.42327943444252014,
|
| 1743 |
+
"learning_rate": 0.0001539863687699274,
|
| 1744 |
+
"loss": 0.607,
|
| 1745 |
+
"step": 24800
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 169.96587030716722,
|
| 1749 |
+
"grad_norm": 0.40863099694252014,
|
| 1750 |
+
"learning_rate": 0.0001516342717224992,
|
| 1751 |
+
"loss": 0.6067,
|
| 1752 |
+
"step": 24900
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 170.64846416382252,
|
| 1756 |
+
"grad_norm": 0.4853709638118744,
|
| 1757 |
+
"learning_rate": 0.00014933143988457237,
|
| 1758 |
+
"loss": 0.6056,
|
| 1759 |
+
"step": 25000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 171.33105802047783,
|
| 1763 |
+
"grad_norm": 0.383801132440567,
|
| 1764 |
+
"learning_rate": 0.00014707815804342907,
|
| 1765 |
+
"loss": 0.6063,
|
| 1766 |
+
"step": 25100
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 172.0136518771331,
|
| 1770 |
+
"grad_norm": 0.4241873621940613,
|
| 1771 |
+
"learning_rate": 0.00014487470485858767,
|
| 1772 |
+
"loss": 0.6058,
|
| 1773 |
+
"step": 25200
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 172.6962457337884,
|
| 1777 |
+
"grad_norm": 0.4262351989746094,
|
| 1778 |
+
"learning_rate": 0.00014272135282734167,
|
| 1779 |
+
"loss": 0.6043,
|
| 1780 |
+
"step": 25300
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 173.37883959044368,
|
| 1784 |
+
"grad_norm": 0.38850197196006775,
|
| 1785 |
+
"learning_rate": 0.00014061836825106022,
|
| 1786 |
+
"loss": 0.606,
|
| 1787 |
+
"step": 25400
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 174.06143344709898,
|
| 1791 |
+
"grad_norm": 0.4169691801071167,
|
| 1792 |
+
"learning_rate": 0.00013856601120225574,
|
| 1793 |
+
"loss": 0.6038,
|
| 1794 |
+
"step": 25500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 174.74402730375428,
|
| 1798 |
+
"grad_norm": 0.43140584230422974,
|
| 1799 |
+
"learning_rate": 0.00013656453549242055,
|
| 1800 |
+
"loss": 0.6043,
|
| 1801 |
+
"step": 25600
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 175.42662116040955,
|
| 1805 |
+
"grad_norm": 0.3729684054851532,
|
| 1806 |
+
"learning_rate": 0.00013461418864063861,
|
| 1807 |
+
"loss": 0.6047,
|
| 1808 |
+
"step": 25700
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 176.10921501706486,
|
| 1812 |
+
"grad_norm": 0.435611367225647,
|
| 1813 |
+
"learning_rate": 0.00013271521184297523,
|
| 1814 |
+
"loss": 0.6035,
|
| 1815 |
+
"step": 25800
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 176.79180887372013,
|
| 1819 |
+
"grad_norm": 0.4359755218029022,
|
| 1820 |
+
"learning_rate": 0.00013086783994264905,
|
| 1821 |
+
"loss": 0.6044,
|
| 1822 |
+
"step": 25900
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 177.47440273037543,
|
| 1826 |
+
"grad_norm": 0.3867483139038086,
|
| 1827 |
+
"learning_rate": 0.00012907230140098895,
|
| 1828 |
+
"loss": 0.6031,
|
| 1829 |
+
"step": 26000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 178.1569965870307,
|
| 1833 |
+
"grad_norm": 0.41799867153167725,
|
| 1834 |
+
"learning_rate": 0.00012732881826918094,
|
| 1835 |
+
"loss": 0.6031,
|
| 1836 |
+
"step": 26100
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 178.839590443686,
|
| 1840 |
+
"grad_norm": 0.43031203746795654,
|
| 1841 |
+
"learning_rate": 0.00012563760616080745,
|
| 1842 |
+
"loss": 0.6028,
|
| 1843 |
+
"step": 26200
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 179.5221843003413,
|
| 1847 |
+
"grad_norm": 0.3603518605232239,
|
| 1848 |
+
"learning_rate": 0.00012399887422518243,
|
| 1849 |
+
"loss": 0.6026,
|
| 1850 |
+
"step": 26300
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 180.20477815699658,
|
| 1854 |
+
"grad_norm": 0.42124807834625244,
|
| 1855 |
+
"learning_rate": 0.0001224128251214867,
|
| 1856 |
+
"loss": 0.6021,
|
| 1857 |
+
"step": 26400
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 180.88737201365188,
|
| 1861 |
+
"grad_norm": 0.40946000814437866,
|
| 1862 |
+
"learning_rate": 0.00012087965499370513,
|
| 1863 |
+
"loss": 0.6036,
|
| 1864 |
+
"step": 26500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 181.56996587030716,
|
| 1868 |
+
"grad_norm": 0.3763185143470764,
|
| 1869 |
+
"learning_rate": 0.00011939955344636983,
|
| 1870 |
+
"loss": 0.6022,
|
| 1871 |
+
"step": 26600
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 182.25255972696246,
|
| 1875 |
+
"grad_norm": 0.4167167544364929,
|
| 1876 |
+
"learning_rate": 0.00011797270352111224,
|
| 1877 |
+
"loss": 0.6016,
|
| 1878 |
+
"step": 26700
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 182.93515358361773,
|
| 1882 |
+
"grad_norm": 0.42429491877555847,
|
| 1883 |
+
"learning_rate": 0.0001165992816740268,
|
| 1884 |
+
"loss": 0.6021,
|
| 1885 |
+
"step": 26800
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 183.61774744027304,
|
| 1889 |
+
"grad_norm": 0.38188785314559937,
|
| 1890 |
+
"learning_rate": 0.00011527945775384831,
|
| 1891 |
+
"loss": 0.6022,
|
| 1892 |
+
"step": 26900
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 184.30034129692834,
|
| 1896 |
+
"grad_norm": 0.47042107582092285,
|
| 1897 |
+
"learning_rate": 0.00011401339498094774,
|
| 1898 |
+
"loss": 0.6016,
|
| 1899 |
+
"step": 27000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 184.9829351535836,
|
| 1903 |
+
"grad_norm": 0.422690212726593,
|
| 1904 |
+
"learning_rate": 0.00011280124992714677,
|
| 1905 |
+
"loss": 0.602,
|
| 1906 |
+
"step": 27100
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 185.6655290102389,
|
| 1910 |
+
"grad_norm": 0.3714023232460022,
|
| 1911 |
+
"learning_rate": 0.00011164317249635475,
|
| 1912 |
+
"loss": 0.6014,
|
| 1913 |
+
"step": 27200
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 186.3481228668942,
|
| 1917 |
+
"grad_norm": 0.43933653831481934,
|
| 1918 |
+
"learning_rate": 0.00011053930590603041,
|
| 1919 |
+
"loss": 0.6003,
|
| 1920 |
+
"step": 27300
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 187.0307167235495,
|
| 1924 |
+
"grad_norm": 0.3583015501499176,
|
| 1925 |
+
"learning_rate": 0.00010948978666947039,
|
| 1926 |
+
"loss": 0.6016,
|
| 1927 |
+
"step": 27400
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 187.7133105802048,
|
| 1931 |
+
"grad_norm": 0.38929620385169983,
|
| 1932 |
+
"learning_rate": 0.00010849474457892682,
|
| 1933 |
+
"loss": 0.6006,
|
| 1934 |
+
"step": 27500
|
| 1935 |
+
}
|
| 1936 |
+
],
|
| 1937 |
+
"logging_steps": 100,
|
| 1938 |
+
"max_steps": 29200,
|
| 1939 |
+
"num_input_tokens_seen": 0,
|
| 1940 |
+
"num_train_epochs": 200,
|
| 1941 |
+
"save_steps": 500,
|
| 1942 |
+
"stateful_callbacks": {
|
| 1943 |
+
"TrainerControl": {
|
| 1944 |
+
"args": {
|
| 1945 |
+
"should_epoch_stop": false,
|
| 1946 |
+
"should_evaluate": false,
|
| 1947 |
+
"should_log": false,
|
| 1948 |
+
"should_save": true,
|
| 1949 |
+
"should_training_stop": false
|
| 1950 |
+
},
|
| 1951 |
+
"attributes": {}
|
| 1952 |
+
}
|
| 1953 |
+
},
|
| 1954 |
+
"total_flos": 8.797169977204408e+17,
|
| 1955 |
+
"train_batch_size": 256,
|
| 1956 |
+
"trial_name": null,
|
| 1957 |
+
"trial_params": null
|
| 1958 |
+
}
|
checkpoint-27500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b3988ed6f3d05a9a0d746b5535dfdb8d9bd5cdf3d3059847445b0b1d44dcf65
|
| 3 |
+
size 5240
|
checkpoint-27500/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-28000/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_function": "gelu_new",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPTJForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attn_pdrop": 0.1,
|
| 7 |
+
"bos_token_id": 50256,
|
| 8 |
+
"embd_pdrop": 0.1,
|
| 9 |
+
"eos_token_id": 50256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"layer_norm_epsilon": 1e-05,
|
| 12 |
+
"model_type": "gptj",
|
| 13 |
+
"n_embd": 768,
|
| 14 |
+
"n_head": 12,
|
| 15 |
+
"n_inner": null,
|
| 16 |
+
"n_layer": 12,
|
| 17 |
+
"n_positions": 1024,
|
| 18 |
+
"resid_pdrop": 0.1,
|
| 19 |
+
"rotary_dim": 64,
|
| 20 |
+
"tie_word_embeddings": false,
|
| 21 |
+
"torch_dtype": "float32",
|
| 22 |
+
"transformers_version": "4.46.2",
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 50257
|
| 25 |
+
}
|
checkpoint-28000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.46.2"
|
| 6 |
+
}
|
checkpoint-28000/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-28000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6298d4cc6f2ccefdc83d1ca809c8d90d4697eff6bf4017ab386dd5ac6ab1cc4c
|
| 3 |
+
size 648995876
|
checkpoint-28000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cb6b3b3d2d2375148064157ae5011aeb2483b6c698c7588409e6ece05ab651a7
|
| 3 |
+
size 1298039738
|
checkpoint-28000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:96a9a3753dcbde9e17820cd06d7a111a8f72dd231d7117ebc1c05d96876cb7bf
|
| 3 |
+
size 14512
|
checkpoint-28000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b73fc8a37bf0ecdc622a9abe6de4b3c4eec9a311ae644a1ff71086bb9bdcf5a7
|
| 3 |
+
size 14512
|
checkpoint-28000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e58b68ea36ad47b639a797350312716f8b56c8f7b7f013ce0f8e0459cbc370b7
|
| 3 |
+
size 1000
|
checkpoint-28000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|endoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endoftext|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "<|endoftext|>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<|endoftext|>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": true,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-28000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"50256": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": true,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"bos_token": "<|endoftext|>",
|
| 15 |
+
"clean_up_tokenization_spaces": false,
|
| 16 |
+
"eos_token": "<|endoftext|>",
|
| 17 |
+
"errors": "replace",
|
| 18 |
+
"model_max_length": 1024,
|
| 19 |
+
"pad_token": "<|endoftext|>",
|
| 20 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 21 |
+
"unk_token": "<|endoftext|>"
|
| 22 |
+
}
|
checkpoint-28000/trainer_state.json
ADDED
|
@@ -0,0 +1,1993 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 191.12627986348122,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 28000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.6825938566552902,
|
| 13 |
+
"grad_norm": 1.5864567756652832,
|
| 14 |
+
"learning_rate": 0.0001,
|
| 15 |
+
"loss": 5.292,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.36518771331058,
|
| 20 |
+
"grad_norm": 0.6435245275497437,
|
| 21 |
+
"learning_rate": 0.0002,
|
| 22 |
+
"loss": 1.3885,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 2.04778156996587,
|
| 27 |
+
"grad_norm": 0.6205620169639587,
|
| 28 |
+
"learning_rate": 0.0003,
|
| 29 |
+
"loss": 1.2326,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 2.73037542662116,
|
| 34 |
+
"grad_norm": 0.668118953704834,
|
| 35 |
+
"learning_rate": 0.0004,
|
| 36 |
+
"loss": 1.234,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 3.4129692832764507,
|
| 41 |
+
"grad_norm": 0.5799328684806824,
|
| 42 |
+
"learning_rate": 0.0005,
|
| 43 |
+
"loss": 1.2338,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 4.09556313993174,
|
| 48 |
+
"grad_norm": 0.5505828261375427,
|
| 49 |
+
"learning_rate": 0.0006,
|
| 50 |
+
"loss": 1.2287,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 4.778156996587031,
|
| 55 |
+
"grad_norm": 0.515738844871521,
|
| 56 |
+
"learning_rate": 0.0007,
|
| 57 |
+
"loss": 1.2327,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 5.460750853242321,
|
| 62 |
+
"grad_norm": 0.509695291519165,
|
| 63 |
+
"learning_rate": 0.0008,
|
| 64 |
+
"loss": 1.2565,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 6.143344709897611,
|
| 69 |
+
"grad_norm": 0.6063594818115234,
|
| 70 |
+
"learning_rate": 0.0009000000000000001,
|
| 71 |
+
"loss": 1.2289,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 6.825938566552901,
|
| 76 |
+
"grad_norm": 0.5587486624717712,
|
| 77 |
+
"learning_rate": 0.001,
|
| 78 |
+
"loss": 1.2334,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 7.508532423208191,
|
| 83 |
+
"grad_norm": 0.5273250341415405,
|
| 84 |
+
"learning_rate": 0.0009999721746341338,
|
| 85 |
+
"loss": 1.2316,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 8.19112627986348,
|
| 90 |
+
"grad_norm": 0.593492329120636,
|
| 91 |
+
"learning_rate": 0.000999888701977651,
|
| 92 |
+
"loss": 1.2279,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 8.873720136518772,
|
| 97 |
+
"grad_norm": 0.44249382615089417,
|
| 98 |
+
"learning_rate": 0.000999749592353472,
|
| 99 |
+
"loss": 1.231,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 9.556313993174061,
|
| 104 |
+
"grad_norm": 0.4120616018772125,
|
| 105 |
+
"learning_rate": 0.0009995548629650466,
|
| 106 |
+
"loss": 1.2282,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 10.238907849829351,
|
| 111 |
+
"grad_norm": 0.5016219019889832,
|
| 112 |
+
"learning_rate": 0.0009993045378942262,
|
| 113 |
+
"loss": 1.2261,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 10.921501706484642,
|
| 118 |
+
"grad_norm": 0.5626952648162842,
|
| 119 |
+
"learning_rate": 0.0009989986480982847,
|
| 120 |
+
"loss": 1.2294,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 11.604095563139932,
|
| 125 |
+
"grad_norm": 0.45088785886764526,
|
| 126 |
+
"learning_rate": 0.000998637231406091,
|
| 127 |
+
"loss": 1.2268,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 12.286689419795222,
|
| 132 |
+
"grad_norm": 0.39686086773872375,
|
| 133 |
+
"learning_rate": 0.0009982203325134304,
|
| 134 |
+
"loss": 1.2254,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 12.969283276450511,
|
| 139 |
+
"grad_norm": 0.3904390037059784,
|
| 140 |
+
"learning_rate": 0.0009977480029774771,
|
| 141 |
+
"loss": 1.2274,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 13.651877133105803,
|
| 146 |
+
"grad_norm": 0.38391488790512085,
|
| 147 |
+
"learning_rate": 0.0009972203012104186,
|
| 148 |
+
"loss": 1.2247,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 14.334470989761092,
|
| 153 |
+
"grad_norm": 0.4345687925815582,
|
| 154 |
+
"learning_rate": 0.0009966372924722312,
|
| 155 |
+
"loss": 1.2252,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 15.017064846416382,
|
| 160 |
+
"grad_norm": 0.34943118691444397,
|
| 161 |
+
"learning_rate": 0.0009959990488626104,
|
| 162 |
+
"loss": 1.2266,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 15.699658703071673,
|
| 167 |
+
"grad_norm": 0.3420363962650299,
|
| 168 |
+
"learning_rate": 0.0009953056493120537,
|
| 169 |
+
"loss": 1.224,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 16.38225255972696,
|
| 174 |
+
"grad_norm": 0.3876005709171295,
|
| 175 |
+
"learning_rate": 0.0009945571795721,
|
| 176 |
+
"loss": 1.2241,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 17.064846416382252,
|
| 181 |
+
"grad_norm": 0.3726520836353302,
|
| 182 |
+
"learning_rate": 0.0009937537322047237,
|
| 183 |
+
"loss": 1.2258,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 17.747440273037544,
|
| 188 |
+
"grad_norm": 0.4071378707885742,
|
| 189 |
+
"learning_rate": 0.0009928954065708895,
|
| 190 |
+
"loss": 1.2235,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 18.43003412969283,
|
| 195 |
+
"grad_norm": 0.443796843290329,
|
| 196 |
+
"learning_rate": 0.0009919823088182628,
|
| 197 |
+
"loss": 1.2346,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 19.112627986348123,
|
| 202 |
+
"grad_norm": 0.3463057279586792,
|
| 203 |
+
"learning_rate": 0.000991014551868084,
|
| 204 |
+
"loss": 1.2292,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 19.795221843003414,
|
| 209 |
+
"grad_norm": 0.31187331676483154,
|
| 210 |
+
"learning_rate": 0.0009899922554012027,
|
| 211 |
+
"loss": 1.2237,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 20.477815699658702,
|
| 216 |
+
"grad_norm": 0.3052064776420593,
|
| 217 |
+
"learning_rate": 0.0009889155458432776,
|
| 218 |
+
"loss": 1.2242,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 21.160409556313994,
|
| 223 |
+
"grad_norm": 0.3126174807548523,
|
| 224 |
+
"learning_rate": 0.000987784556349142,
|
| 225 |
+
"loss": 1.2251,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 21.843003412969285,
|
| 230 |
+
"grad_norm": 0.29788631200790405,
|
| 231 |
+
"learning_rate": 0.0009865994267863351,
|
| 232 |
+
"loss": 1.2224,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 22.525597269624573,
|
| 237 |
+
"grad_norm": 0.2966502606868744,
|
| 238 |
+
"learning_rate": 0.000985360303717807,
|
| 239 |
+
"loss": 1.2233,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 23.208191126279864,
|
| 244 |
+
"grad_norm": 0.31454625725746155,
|
| 245 |
+
"learning_rate": 0.0009840673403837918,
|
| 246 |
+
"loss": 1.2242,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 23.890784982935152,
|
| 251 |
+
"grad_norm": 0.33144181966781616,
|
| 252 |
+
"learning_rate": 0.0009827206966828582,
|
| 253 |
+
"loss": 1.2219,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 24.573378839590443,
|
| 258 |
+
"grad_norm": 0.3603832423686981,
|
| 259 |
+
"learning_rate": 0.000981320539152133,
|
| 260 |
+
"loss": 1.2227,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 25.255972696245735,
|
| 265 |
+
"grad_norm": 0.28107529878616333,
|
| 266 |
+
"learning_rate": 0.0009798670409467075,
|
| 267 |
+
"loss": 1.2236,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 25.938566552901023,
|
| 272 |
+
"grad_norm": 0.3241814076900482,
|
| 273 |
+
"learning_rate": 0.0009783603818182237,
|
| 274 |
+
"loss": 1.2212,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 26.621160409556314,
|
| 279 |
+
"grad_norm": 0.35018429160118103,
|
| 280 |
+
"learning_rate": 0.0009768007480926437,
|
| 281 |
+
"loss": 1.2223,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 27.303754266211605,
|
| 286 |
+
"grad_norm": 0.2918601632118225,
|
| 287 |
+
"learning_rate": 0.0009751883326472076,
|
| 288 |
+
"loss": 1.2228,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 27.986348122866893,
|
| 293 |
+
"grad_norm": 0.29299041628837585,
|
| 294 |
+
"learning_rate": 0.0009735233348865806,
|
| 295 |
+
"loss": 1.2208,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 28.668941979522184,
|
| 300 |
+
"grad_norm": 0.3291133642196655,
|
| 301 |
+
"learning_rate": 0.0009718059607181931,
|
| 302 |
+
"loss": 1.2219,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 29.351535836177476,
|
| 307 |
+
"grad_norm": 0.2924804985523224,
|
| 308 |
+
"learning_rate": 0.0009700364225267769,
|
| 309 |
+
"loss": 1.2222,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 30.034129692832764,
|
| 314 |
+
"grad_norm": 0.3081340193748474,
|
| 315 |
+
"learning_rate": 0.0009682149391480986,
|
| 316 |
+
"loss": 1.2201,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 30.716723549488055,
|
| 321 |
+
"grad_norm": 0.2960987091064453,
|
| 322 |
+
"learning_rate": 0.0009663417358418981,
|
| 323 |
+
"loss": 1.2213,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 31.399317406143346,
|
| 328 |
+
"grad_norm": 0.29421886801719666,
|
| 329 |
+
"learning_rate": 0.0009644170442640303,
|
| 330 |
+
"loss": 1.2201,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 32.081911262798634,
|
| 335 |
+
"grad_norm": 0.3381507098674774,
|
| 336 |
+
"learning_rate": 0.0009624411024378166,
|
| 337 |
+
"loss": 1.2186,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 32.76450511945392,
|
| 342 |
+
"grad_norm": 0.3033551275730133,
|
| 343 |
+
"learning_rate": 0.0009604141547246092,
|
| 344 |
+
"loss": 1.2199,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 33.44709897610922,
|
| 349 |
+
"grad_norm": 0.3156335949897766,
|
| 350 |
+
"learning_rate": 0.0009583364517935713,
|
| 351 |
+
"loss": 1.2189,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 34.129692832764505,
|
| 356 |
+
"grad_norm": 0.33952537178993225,
|
| 357 |
+
"learning_rate": 0.0009562082505906771,
|
| 358 |
+
"loss": 1.2163,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 34.81228668941979,
|
| 363 |
+
"grad_norm": 0.2941412925720215,
|
| 364 |
+
"learning_rate": 0.0009540298143069361,
|
| 365 |
+
"loss": 1.2179,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 35.49488054607509,
|
| 370 |
+
"grad_norm": 0.28668472170829773,
|
| 371 |
+
"learning_rate": 0.0009518014123458442,
|
| 372 |
+
"loss": 1.2158,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 36.177474402730375,
|
| 377 |
+
"grad_norm": 0.2789587080478668,
|
| 378 |
+
"learning_rate": 0.0009495233202900674,
|
| 379 |
+
"loss": 1.213,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 36.86006825938566,
|
| 384 |
+
"grad_norm": 0.3137347102165222,
|
| 385 |
+
"learning_rate": 0.0009471958198673614,
|
| 386 |
+
"loss": 1.214,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 37.54266211604096,
|
| 391 |
+
"grad_norm": 0.25685933232307434,
|
| 392 |
+
"learning_rate": 0.0009448191989157294,
|
| 393 |
+
"loss": 1.21,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 38.225255972696246,
|
| 398 |
+
"grad_norm": 0.3252381980419159,
|
| 399 |
+
"learning_rate": 0.0009423937513478271,
|
| 400 |
+
"loss": 1.2071,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 38.907849829351534,
|
| 405 |
+
"grad_norm": 0.2932053208351135,
|
| 406 |
+
"learning_rate": 0.0009399197771146144,
|
| 407 |
+
"loss": 1.2077,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 39.59044368600683,
|
| 412 |
+
"grad_norm": 0.33668607473373413,
|
| 413 |
+
"learning_rate": 0.0009373975821682607,
|
| 414 |
+
"loss": 1.2019,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 40.27303754266212,
|
| 419 |
+
"grad_norm": 0.31952792406082153,
|
| 420 |
+
"learning_rate": 0.0009348274784243094,
|
| 421 |
+
"loss": 1.1979,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 40.955631399317404,
|
| 426 |
+
"grad_norm": 0.3525752127170563,
|
| 427 |
+
"learning_rate": 0.0009322097837231025,
|
| 428 |
+
"loss": 1.1999,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 41.6382252559727,
|
| 433 |
+
"grad_norm": 0.3389936685562134,
|
| 434 |
+
"learning_rate": 0.0009295448217904745,
|
| 435 |
+
"loss": 1.1923,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 42.32081911262799,
|
| 440 |
+
"grad_norm": 0.3114762008190155,
|
| 441 |
+
"learning_rate": 0.0009268329221977179,
|
| 442 |
+
"loss": 1.19,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 43.003412969283275,
|
| 447 |
+
"grad_norm": 0.3166259229183197,
|
| 448 |
+
"learning_rate": 0.0009240744203208256,
|
| 449 |
+
"loss": 1.1914,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 43.68600682593857,
|
| 454 |
+
"grad_norm": 0.31299668550491333,
|
| 455 |
+
"learning_rate": 0.0009212696572990148,
|
| 456 |
+
"loss": 1.1843,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 44.36860068259386,
|
| 461 |
+
"grad_norm": 0.36728623509407043,
|
| 462 |
+
"learning_rate": 0.0009184189799925403,
|
| 463 |
+
"loss": 1.1828,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 45.051194539249146,
|
| 468 |
+
"grad_norm": 0.25386592745780945,
|
| 469 |
+
"learning_rate": 0.0009155227409397977,
|
| 470 |
+
"loss": 1.1849,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 45.73378839590444,
|
| 475 |
+
"grad_norm": 0.3170667290687561,
|
| 476 |
+
"learning_rate": 0.0009125812983137259,
|
| 477 |
+
"loss": 1.1789,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 46.41638225255973,
|
| 482 |
+
"grad_norm": 0.3354248106479645,
|
| 483 |
+
"learning_rate": 0.0009095950158775128,
|
| 484 |
+
"loss": 1.1781,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 47.098976109215016,
|
| 489 |
+
"grad_norm": 0.2761414647102356,
|
| 490 |
+
"learning_rate": 0.0009065642629396087,
|
| 491 |
+
"loss": 1.1803,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 47.781569965870304,
|
| 496 |
+
"grad_norm": 0.28061142563819885,
|
| 497 |
+
"learning_rate": 0.0009034894143080554,
|
| 498 |
+
"loss": 1.175,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 48.4641638225256,
|
| 503 |
+
"grad_norm": 0.3548315763473511,
|
| 504 |
+
"learning_rate": 0.0009003708502441333,
|
| 505 |
+
"loss": 1.1743,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 49.14675767918089,
|
| 510 |
+
"grad_norm": 0.3077629804611206,
|
| 511 |
+
"learning_rate": 0.0008972089564153364,
|
| 512 |
+
"loss": 1.1762,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 49.829351535836174,
|
| 517 |
+
"grad_norm": 0.3600843846797943,
|
| 518 |
+
"learning_rate": 0.0008940041238476759,
|
| 519 |
+
"loss": 1.1718,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 50.51194539249147,
|
| 524 |
+
"grad_norm": 0.3295997083187103,
|
| 525 |
+
"learning_rate": 0.0008907567488773241,
|
| 526 |
+
"loss": 1.1709,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 51.19453924914676,
|
| 531 |
+
"grad_norm": 0.32083791494369507,
|
| 532 |
+
"learning_rate": 0.0008874672331015996,
|
| 533 |
+
"loss": 1.1727,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 51.877133105802045,
|
| 538 |
+
"grad_norm": 0.32850903272628784,
|
| 539 |
+
"learning_rate": 0.0008841359833293022,
|
| 540 |
+
"loss": 1.1685,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 52.55972696245734,
|
| 545 |
+
"grad_norm": 0.33856746554374695,
|
| 546 |
+
"learning_rate": 0.000880763411530404,
|
| 547 |
+
"loss": 1.168,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 53.24232081911263,
|
| 552 |
+
"grad_norm": 0.34316346049308777,
|
| 553 |
+
"learning_rate": 0.0008773499347851014,
|
| 554 |
+
"loss": 1.1687,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 53.924914675767916,
|
| 559 |
+
"grad_norm": 0.38703614473342896,
|
| 560 |
+
"learning_rate": 0.0008738959752322358,
|
| 561 |
+
"loss": 1.165,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 54.60750853242321,
|
| 566 |
+
"grad_norm": 0.3954058587551117,
|
| 567 |
+
"learning_rate": 0.0008704019600170884,
|
| 568 |
+
"loss": 1.1637,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 55.2901023890785,
|
| 573 |
+
"grad_norm": 0.33417293429374695,
|
| 574 |
+
"learning_rate": 0.0008668683212385556,
|
| 575 |
+
"loss": 1.1635,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 55.972696245733786,
|
| 580 |
+
"grad_norm": 0.3305743932723999,
|
| 581 |
+
"learning_rate": 0.0008632954958957123,
|
| 582 |
+
"loss": 1.1601,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 56.65529010238908,
|
| 587 |
+
"grad_norm": 0.3755279779434204,
|
| 588 |
+
"learning_rate": 0.0008596839258337689,
|
| 589 |
+
"loss": 1.158,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 57.33788395904437,
|
| 594 |
+
"grad_norm": 0.3553084433078766,
|
| 595 |
+
"learning_rate": 0.000856034057689429,
|
| 596 |
+
"loss": 1.1568,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 58.02047781569966,
|
| 601 |
+
"grad_norm": 0.38361021876335144,
|
| 602 |
+
"learning_rate": 0.0008523463428356549,
|
| 603 |
+
"loss": 1.154,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 58.70307167235495,
|
| 608 |
+
"grad_norm": 0.35759079456329346,
|
| 609 |
+
"learning_rate": 0.0008486212373258465,
|
| 610 |
+
"loss": 1.152,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 59.38566552901024,
|
| 615 |
+
"grad_norm": 0.32452645897865295,
|
| 616 |
+
"learning_rate": 0.0008448592018374424,
|
| 617 |
+
"loss": 1.1497,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 60.06825938566553,
|
| 622 |
+
"grad_norm": 0.37652239203453064,
|
| 623 |
+
"learning_rate": 0.0008410607016149491,
|
| 624 |
+
"loss": 1.1461,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 60.75085324232082,
|
| 629 |
+
"grad_norm": 0.35698965191841125,
|
| 630 |
+
"learning_rate": 0.0008372262064124035,
|
| 631 |
+
"loss": 1.1453,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 61.43344709897611,
|
| 636 |
+
"grad_norm": 0.3505449891090393,
|
| 637 |
+
"learning_rate": 0.0008333561904352813,
|
| 638 |
+
"loss": 1.1409,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 62.1160409556314,
|
| 643 |
+
"grad_norm": 0.3646785616874695,
|
| 644 |
+
"learning_rate": 0.000829451132281851,
|
| 645 |
+
"loss": 1.1363,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 62.79863481228669,
|
| 650 |
+
"grad_norm": 0.34945082664489746,
|
| 651 |
+
"learning_rate": 0.0008255115148839877,
|
| 652 |
+
"loss": 1.1371,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 63.48122866894198,
|
| 657 |
+
"grad_norm": 0.34238216280937195,
|
| 658 |
+
"learning_rate": 0.0008215378254474485,
|
| 659 |
+
"loss": 1.1296,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 64.16382252559727,
|
| 664 |
+
"grad_norm": 0.37401944398880005,
|
| 665 |
+
"learning_rate": 0.0008175305553916226,
|
| 666 |
+
"loss": 1.1241,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 64.84641638225256,
|
| 671 |
+
"grad_norm": 0.3763269782066345,
|
| 672 |
+
"learning_rate": 0.000813490200288756,
|
| 673 |
+
"loss": 1.1257,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 65.52901023890784,
|
| 678 |
+
"grad_norm": 0.3553602993488312,
|
| 679 |
+
"learning_rate": 0.0008094172598026662,
|
| 680 |
+
"loss": 1.1147,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 66.21160409556315,
|
| 685 |
+
"grad_norm": 0.36700302362442017,
|
| 686 |
+
"learning_rate": 0.0008053122376269497,
|
| 687 |
+
"loss": 1.1078,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 66.89419795221843,
|
| 692 |
+
"grad_norm": 0.4266979694366455,
|
| 693 |
+
"learning_rate": 0.0008011756414226908,
|
| 694 |
+
"loss": 1.1114,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 67.57679180887372,
|
| 699 |
+
"grad_norm": 0.3603265583515167,
|
| 700 |
+
"learning_rate": 0.0007970079827556795,
|
| 701 |
+
"loss": 1.0967,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 68.25938566552901,
|
| 706 |
+
"grad_norm": 0.39329585433006287,
|
| 707 |
+
"learning_rate": 0.000792809777033148,
|
| 708 |
+
"loss": 1.0898,
|
| 709 |
+
"step": 10000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 68.9419795221843,
|
| 713 |
+
"grad_norm": 0.40609198808670044,
|
| 714 |
+
"learning_rate": 0.0007885815434400306,
|
| 715 |
+
"loss": 1.0944,
|
| 716 |
+
"step": 10100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 69.62457337883959,
|
| 720 |
+
"grad_norm": 0.3622093200683594,
|
| 721 |
+
"learning_rate": 0.0007843238048747572,
|
| 722 |
+
"loss": 1.0761,
|
| 723 |
+
"step": 10200
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 70.30716723549489,
|
| 727 |
+
"grad_norm": 0.39394432306289673,
|
| 728 |
+
"learning_rate": 0.0007800370878845866,
|
| 729 |
+
"loss": 1.0692,
|
| 730 |
+
"step": 10300
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 70.98976109215018,
|
| 734 |
+
"grad_norm": 0.382322758436203,
|
| 735 |
+
"learning_rate": 0.0007757219226004912,
|
| 736 |
+
"loss": 1.0767,
|
| 737 |
+
"step": 10400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 71.67235494880546,
|
| 741 |
+
"grad_norm": 0.3804159462451935,
|
| 742 |
+
"learning_rate": 0.0007713788426715937,
|
| 743 |
+
"loss": 1.0543,
|
| 744 |
+
"step": 10500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 72.35494880546075,
|
| 748 |
+
"grad_norm": 0.3960065245628357,
|
| 749 |
+
"learning_rate": 0.0007670083851991749,
|
| 750 |
+
"loss": 1.0477,
|
| 751 |
+
"step": 10600
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 73.03754266211604,
|
| 755 |
+
"grad_norm": 0.39962372183799744,
|
| 756 |
+
"learning_rate": 0.0007626110906702484,
|
| 757 |
+
"loss": 1.0561,
|
| 758 |
+
"step": 10700
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 73.72013651877133,
|
| 762 |
+
"grad_norm": 0.4019100069999695,
|
| 763 |
+
"learning_rate": 0.0007581875028907215,
|
| 764 |
+
"loss": 1.0329,
|
| 765 |
+
"step": 10800
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 74.40273037542661,
|
| 769 |
+
"grad_norm": 0.42911744117736816,
|
| 770 |
+
"learning_rate": 0.0007537381689181425,
|
| 771 |
+
"loss": 1.0265,
|
| 772 |
+
"step": 10900
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 75.08532423208192,
|
| 776 |
+
"grad_norm": 0.4002211093902588,
|
| 777 |
+
"learning_rate": 0.0007492636389940478,
|
| 778 |
+
"loss": 1.0332,
|
| 779 |
+
"step": 11000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 75.7679180887372,
|
| 783 |
+
"grad_norm": 0.40011659264564514,
|
| 784 |
+
"learning_rate": 0.0007447644664759138,
|
| 785 |
+
"loss": 1.0127,
|
| 786 |
+
"step": 11100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 76.45051194539249,
|
| 790 |
+
"grad_norm": 0.43485236167907715,
|
| 791 |
+
"learning_rate": 0.0007402412077687246,
|
| 792 |
+
"loss": 1.0047,
|
| 793 |
+
"step": 11200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 77.13310580204778,
|
| 797 |
+
"grad_norm": 0.4095554053783417,
|
| 798 |
+
"learning_rate": 0.0007356944222561622,
|
| 799 |
+
"loss": 1.0114,
|
| 800 |
+
"step": 11300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 77.81569965870307,
|
| 804 |
+
"grad_norm": 0.39631539583206177,
|
| 805 |
+
"learning_rate": 0.0007311246722314281,
|
| 806 |
+
"loss": 0.9916,
|
| 807 |
+
"step": 11400
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 78.49829351535836,
|
| 811 |
+
"grad_norm": 0.4598711133003235,
|
| 812 |
+
"learning_rate": 0.0007265325228277065,
|
| 813 |
+
"loss": 0.9822,
|
| 814 |
+
"step": 11500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 79.18088737201366,
|
| 818 |
+
"grad_norm": 0.39505425095558167,
|
| 819 |
+
"learning_rate": 0.0007219185419482739,
|
| 820 |
+
"loss": 0.9883,
|
| 821 |
+
"step": 11600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 79.86348122866895,
|
| 825 |
+
"grad_norm": 0.3844675123691559,
|
| 826 |
+
"learning_rate": 0.0007172833001962683,
|
| 827 |
+
"loss": 0.9701,
|
| 828 |
+
"step": 11700
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 80.54607508532423,
|
| 832 |
+
"grad_norm": 0.48971033096313477,
|
| 833 |
+
"learning_rate": 0.000712627370804124,
|
| 834 |
+
"loss": 0.9597,
|
| 835 |
+
"step": 11800
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 81.22866894197952,
|
| 839 |
+
"grad_norm": 0.40114977955818176,
|
| 840 |
+
"learning_rate": 0.0007079513295626796,
|
| 841 |
+
"loss": 0.9643,
|
| 842 |
+
"step": 11900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 81.91126279863481,
|
| 846 |
+
"grad_norm": 0.42664656043052673,
|
| 847 |
+
"learning_rate": 0.0007032557547499727,
|
| 848 |
+
"loss": 0.9482,
|
| 849 |
+
"step": 12000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 82.5938566552901,
|
| 853 |
+
"grad_norm": 0.5133378505706787,
|
| 854 |
+
"learning_rate": 0.0006985412270597231,
|
| 855 |
+
"loss": 0.9363,
|
| 856 |
+
"step": 12100
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 83.2764505119454,
|
| 860 |
+
"grad_norm": 0.43057283759117126,
|
| 861 |
+
"learning_rate": 0.000693808329529522,
|
| 862 |
+
"loss": 0.9409,
|
| 863 |
+
"step": 12200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 83.95904436860069,
|
| 867 |
+
"grad_norm": 0.4190930426120758,
|
| 868 |
+
"learning_rate": 0.0006890576474687264,
|
| 869 |
+
"loss": 0.9255,
|
| 870 |
+
"step": 12300
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 84.64163822525597,
|
| 874 |
+
"grad_norm": 0.43727907538414,
|
| 875 |
+
"learning_rate": 0.000684289768386076,
|
| 876 |
+
"loss": 0.9119,
|
| 877 |
+
"step": 12400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 85.32423208191126,
|
| 881 |
+
"grad_norm": 0.4994739890098572,
|
| 882 |
+
"learning_rate": 0.000679505281917037,
|
| 883 |
+
"loss": 0.914,
|
| 884 |
+
"step": 12500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 86.00682593856655,
|
| 888 |
+
"grad_norm": 0.45477139949798584,
|
| 889 |
+
"learning_rate": 0.0006747047797508828,
|
| 890 |
+
"loss": 0.9017,
|
| 891 |
+
"step": 12600
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 86.68941979522184,
|
| 895 |
+
"grad_norm": 0.4953708052635193,
|
| 896 |
+
"learning_rate": 0.0006698888555575204,
|
| 897 |
+
"loss": 0.8881,
|
| 898 |
+
"step": 12700
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 87.37201365187714,
|
| 902 |
+
"grad_norm": 0.4449358284473419,
|
| 903 |
+
"learning_rate": 0.000665058104914073,
|
| 904 |
+
"loss": 0.8875,
|
| 905 |
+
"step": 12800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 88.05460750853243,
|
| 909 |
+
"grad_norm": 0.4819263815879822,
|
| 910 |
+
"learning_rate": 0.0006602131252312249,
|
| 911 |
+
"loss": 0.8748,
|
| 912 |
+
"step": 12900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 88.73720136518772,
|
| 916 |
+
"grad_norm": 0.5412826538085938,
|
| 917 |
+
"learning_rate": 0.0006553545156793422,
|
| 918 |
+
"loss": 0.8655,
|
| 919 |
+
"step": 13000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 89.419795221843,
|
| 923 |
+
"grad_norm": 0.4753580391407013,
|
| 924 |
+
"learning_rate": 0.0006504828771143733,
|
| 925 |
+
"loss": 0.8617,
|
| 926 |
+
"step": 13100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 90.10238907849829,
|
| 930 |
+
"grad_norm": 0.545566737651825,
|
| 931 |
+
"learning_rate": 0.0006455988120035418,
|
| 932 |
+
"loss": 0.85,
|
| 933 |
+
"step": 13200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 90.78498293515358,
|
| 937 |
+
"grad_norm": 0.5255779027938843,
|
| 938 |
+
"learning_rate": 0.0006407029243508423,
|
| 939 |
+
"loss": 0.8449,
|
| 940 |
+
"step": 13300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 91.46757679180887,
|
| 944 |
+
"grad_norm": 0.43862029910087585,
|
| 945 |
+
"learning_rate": 0.0006357958196223419,
|
| 946 |
+
"loss": 0.8371,
|
| 947 |
+
"step": 13400
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 92.15017064846417,
|
| 951 |
+
"grad_norm": 0.5360426902770996,
|
| 952 |
+
"learning_rate": 0.0006308781046713046,
|
| 953 |
+
"loss": 0.8257,
|
| 954 |
+
"step": 13500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 92.83276450511946,
|
| 958 |
+
"grad_norm": 0.5451890230178833,
|
| 959 |
+
"learning_rate": 0.0006259503876631428,
|
| 960 |
+
"loss": 0.8247,
|
| 961 |
+
"step": 13600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 93.51535836177474,
|
| 965 |
+
"grad_norm": 0.4524667263031006,
|
| 966 |
+
"learning_rate": 0.0006210132780002057,
|
| 967 |
+
"loss": 0.8147,
|
| 968 |
+
"step": 13700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 94.19795221843003,
|
| 972 |
+
"grad_norm": 0.6203665733337402,
|
| 973 |
+
"learning_rate": 0.0006160673862464167,
|
| 974 |
+
"loss": 0.8022,
|
| 975 |
+
"step": 13800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 94.88054607508532,
|
| 979 |
+
"grad_norm": 0.5576283931732178,
|
| 980 |
+
"learning_rate": 0.0006111133240517648,
|
| 981 |
+
"loss": 0.8057,
|
| 982 |
+
"step": 13900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 95.56313993174061,
|
| 986 |
+
"grad_norm": 0.48380613327026367,
|
| 987 |
+
"learning_rate": 0.000606151704076664,
|
| 988 |
+
"loss": 0.7906,
|
| 989 |
+
"step": 14000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 96.24573378839591,
|
| 993 |
+
"grad_norm": 0.5328605771064758,
|
| 994 |
+
"learning_rate": 0.0006011831399161856,
|
| 995 |
+
"loss": 0.7823,
|
| 996 |
+
"step": 14100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 96.9283276450512,
|
| 1000 |
+
"grad_norm": 0.5395786762237549,
|
| 1001 |
+
"learning_rate": 0.0005962082460241767,
|
| 1002 |
+
"loss": 0.7883,
|
| 1003 |
+
"step": 14200
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 97.61092150170649,
|
| 1007 |
+
"grad_norm": 0.48329558968544006,
|
| 1008 |
+
"learning_rate": 0.0005912276376372715,
|
| 1009 |
+
"loss": 0.7707,
|
| 1010 |
+
"step": 14300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 98.29351535836177,
|
| 1014 |
+
"grad_norm": 0.5496215224266052,
|
| 1015 |
+
"learning_rate": 0.0005862419306988061,
|
| 1016 |
+
"loss": 0.7636,
|
| 1017 |
+
"step": 14400
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 98.97610921501706,
|
| 1021 |
+
"grad_norm": 0.6007915139198303,
|
| 1022 |
+
"learning_rate": 0.0005812517417826456,
|
| 1023 |
+
"loss": 0.7712,
|
| 1024 |
+
"step": 14500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 99.65870307167235,
|
| 1028 |
+
"grad_norm": 0.4981062710285187,
|
| 1029 |
+
"learning_rate": 0.000576257688016934,
|
| 1030 |
+
"loss": 0.7529,
|
| 1031 |
+
"step": 14600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 100.34129692832765,
|
| 1035 |
+
"grad_norm": 0.5475151538848877,
|
| 1036 |
+
"learning_rate": 0.0005712603870077736,
|
| 1037 |
+
"loss": 0.7468,
|
| 1038 |
+
"step": 14700
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 101.02389078498294,
|
| 1042 |
+
"grad_norm": 0.4879900813102722,
|
| 1043 |
+
"learning_rate": 0.0005662604567628494,
|
| 1044 |
+
"loss": 0.7568,
|
| 1045 |
+
"step": 14800
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 101.70648464163823,
|
| 1049 |
+
"grad_norm": 0.4963570535182953,
|
| 1050 |
+
"learning_rate": 0.0005612585156149982,
|
| 1051 |
+
"loss": 0.7386,
|
| 1052 |
+
"step": 14900
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 102.38907849829351,
|
| 1056 |
+
"grad_norm": 0.5414558053016663,
|
| 1057 |
+
"learning_rate": 0.0005562551821457418,
|
| 1058 |
+
"loss": 0.7326,
|
| 1059 |
+
"step": 15000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 103.0716723549488,
|
| 1063 |
+
"grad_norm": 0.47427278757095337,
|
| 1064 |
+
"learning_rate": 0.0005512510751087889,
|
| 1065 |
+
"loss": 0.7415,
|
| 1066 |
+
"step": 15100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 103.75426621160409,
|
| 1070 |
+
"grad_norm": 0.48589903116226196,
|
| 1071 |
+
"learning_rate": 0.0005462468133535131,
|
| 1072 |
+
"loss": 0.7267,
|
| 1073 |
+
"step": 15200
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 104.43686006825939,
|
| 1077 |
+
"grad_norm": 0.5541400909423828,
|
| 1078 |
+
"learning_rate": 0.0005412430157484224,
|
| 1079 |
+
"loss": 0.7206,
|
| 1080 |
+
"step": 15300
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 105.11945392491468,
|
| 1084 |
+
"grad_norm": 0.4941166341304779,
|
| 1085 |
+
"learning_rate": 0.0005362403011046234,
|
| 1086 |
+
"loss": 0.7289,
|
| 1087 |
+
"step": 15400
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 105.80204778156997,
|
| 1091 |
+
"grad_norm": 0.4855073094367981,
|
| 1092 |
+
"learning_rate": 0.0005312392880992954,
|
| 1093 |
+
"loss": 0.716,
|
| 1094 |
+
"step": 15500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 106.48464163822526,
|
| 1098 |
+
"grad_norm": 0.5357630252838135,
|
| 1099 |
+
"learning_rate": 0.0005262405951991788,
|
| 1100 |
+
"loss": 0.7111,
|
| 1101 |
+
"step": 15600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 107.16723549488054,
|
| 1105 |
+
"grad_norm": 0.5147479772567749,
|
| 1106 |
+
"learning_rate": 0.0005212448405840905,
|
| 1107 |
+
"loss": 0.7181,
|
| 1108 |
+
"step": 15700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 107.84982935153583,
|
| 1112 |
+
"grad_norm": 0.47526028752326965,
|
| 1113 |
+
"learning_rate": 0.0005162526420704744,
|
| 1114 |
+
"loss": 0.7072,
|
| 1115 |
+
"step": 15800
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 108.53242320819113,
|
| 1119 |
+
"grad_norm": 0.528582751750946,
|
| 1120 |
+
"learning_rate": 0.0005112646170349981,
|
| 1121 |
+
"loss": 0.7021,
|
| 1122 |
+
"step": 15900
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 109.21501706484642,
|
| 1126 |
+
"grad_norm": 0.4804699420928955,
|
| 1127 |
+
"learning_rate": 0.0005062813823382014,
|
| 1128 |
+
"loss": 0.7093,
|
| 1129 |
+
"step": 16000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 109.89761092150171,
|
| 1133 |
+
"grad_norm": 0.5126569271087646,
|
| 1134 |
+
"learning_rate": 0.000501303554248212,
|
| 1135 |
+
"loss": 0.7008,
|
| 1136 |
+
"step": 16100
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 110.580204778157,
|
| 1140 |
+
"grad_norm": 0.5352160930633545,
|
| 1141 |
+
"learning_rate": 0.0004963317483645311,
|
| 1142 |
+
"loss": 0.6946,
|
| 1143 |
+
"step": 16200
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 111.26279863481228,
|
| 1147 |
+
"grad_norm": 0.5465342402458191,
|
| 1148 |
+
"learning_rate": 0.0004913665795419041,
|
| 1149 |
+
"loss": 0.7008,
|
| 1150 |
+
"step": 16300
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 111.94539249146757,
|
| 1154 |
+
"grad_norm": 0.45367056131362915,
|
| 1155 |
+
"learning_rate": 0.00048640866181428265,
|
| 1156 |
+
"loss": 0.6939,
|
| 1157 |
+
"step": 16400
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 112.62798634812286,
|
| 1161 |
+
"grad_norm": 0.6308451294898987,
|
| 1162 |
+
"learning_rate": 0.00048145860831888776,
|
| 1163 |
+
"loss": 0.6885,
|
| 1164 |
+
"step": 16500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 113.31058020477816,
|
| 1168 |
+
"grad_norm": 0.4779800772666931,
|
| 1169 |
+
"learning_rate": 0.00047651703122038395,
|
| 1170 |
+
"loss": 0.6939,
|
| 1171 |
+
"step": 16600
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 113.99317406143345,
|
| 1175 |
+
"grad_norm": 0.4980040192604065,
|
| 1176 |
+
"learning_rate": 0.0004715845416351744,
|
| 1177 |
+
"loss": 0.6879,
|
| 1178 |
+
"step": 16700
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 114.67576791808874,
|
| 1182 |
+
"grad_norm": 0.5260293483734131,
|
| 1183 |
+
"learning_rate": 0.00046666174955582507,
|
| 1184 |
+
"loss": 0.6835,
|
| 1185 |
+
"step": 16800
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 115.35836177474403,
|
| 1189 |
+
"grad_norm": 0.47831013798713684,
|
| 1190 |
+
"learning_rate": 0.0004617492637756282,
|
| 1191 |
+
"loss": 0.6877,
|
| 1192 |
+
"step": 16900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 116.04095563139931,
|
| 1196 |
+
"grad_norm": 0.5765718221664429,
|
| 1197 |
+
"learning_rate": 0.0004568476918133129,
|
| 1198 |
+
"loss": 0.6813,
|
| 1199 |
+
"step": 17000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 116.7235494880546,
|
| 1203 |
+
"grad_norm": 0.5322794318199158,
|
| 1204 |
+
"learning_rate": 0.00045195763983791573,
|
| 1205 |
+
"loss": 0.679,
|
| 1206 |
+
"step": 17100
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 117.4061433447099,
|
| 1210 |
+
"grad_norm": 0.514970064163208,
|
| 1211 |
+
"learning_rate": 0.00044707971259381584,
|
| 1212 |
+
"loss": 0.68,
|
| 1213 |
+
"step": 17200
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 118.08873720136519,
|
| 1217 |
+
"grad_norm": 0.539171576499939,
|
| 1218 |
+
"learning_rate": 0.00044221451332594785,
|
| 1219 |
+
"loss": 0.6769,
|
| 1220 |
+
"step": 17300
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 118.77133105802048,
|
| 1224 |
+
"grad_norm": 0.5568128824234009,
|
| 1225 |
+
"learning_rate": 0.00043736264370519863,
|
| 1226 |
+
"loss": 0.6757,
|
| 1227 |
+
"step": 17400
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 119.45392491467577,
|
| 1231 |
+
"grad_norm": 0.4861965477466583,
|
| 1232 |
+
"learning_rate": 0.0004325247037540011,
|
| 1233 |
+
"loss": 0.6749,
|
| 1234 |
+
"step": 17500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 120.13651877133105,
|
| 1238 |
+
"grad_norm": 0.49631592631340027,
|
| 1239 |
+
"learning_rate": 0.00042770129177212914,
|
| 1240 |
+
"loss": 0.671,
|
| 1241 |
+
"step": 17600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 120.81911262798634,
|
| 1245 |
+
"grad_norm": 0.5209415555000305,
|
| 1246 |
+
"learning_rate": 0.00042289300426270806,
|
| 1247 |
+
"loss": 0.6714,
|
| 1248 |
+
"step": 17700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 121.50170648464164,
|
| 1252 |
+
"grad_norm": 0.4455106556415558,
|
| 1253 |
+
"learning_rate": 0.0004181004358584447,
|
| 1254 |
+
"loss": 0.6699,
|
| 1255 |
+
"step": 17800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 122.18430034129693,
|
| 1259 |
+
"grad_norm": 0.46950340270996094,
|
| 1260 |
+
"learning_rate": 0.00041332417924809175,
|
| 1261 |
+
"loss": 0.6654,
|
| 1262 |
+
"step": 17900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 122.86689419795222,
|
| 1266 |
+
"grad_norm": 0.5346481800079346,
|
| 1267 |
+
"learning_rate": 0.00040856482510314985,
|
| 1268 |
+
"loss": 0.6681,
|
| 1269 |
+
"step": 18000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 123.54948805460751,
|
| 1273 |
+
"grad_norm": 0.47252893447875977,
|
| 1274 |
+
"learning_rate": 0.0004038229620048207,
|
| 1275 |
+
"loss": 0.6648,
|
| 1276 |
+
"step": 18100
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 124.2320819112628,
|
| 1280 |
+
"grad_norm": 0.5372588634490967,
|
| 1281 |
+
"learning_rate": 0.00039909917637121796,
|
| 1282 |
+
"loss": 0.6611,
|
| 1283 |
+
"step": 18200
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 124.91467576791808,
|
| 1287 |
+
"grad_norm": 0.5461378693580627,
|
| 1288 |
+
"learning_rate": 0.00039439405238484614,
|
| 1289 |
+
"loss": 0.6637,
|
| 1290 |
+
"step": 18300
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 125.59726962457339,
|
| 1294 |
+
"grad_norm": 0.4744736850261688,
|
| 1295 |
+
"learning_rate": 0.0003897081719203558,
|
| 1296 |
+
"loss": 0.659,
|
| 1297 |
+
"step": 18400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 126.27986348122867,
|
| 1301 |
+
"grad_norm": 0.5168038606643677,
|
| 1302 |
+
"learning_rate": 0.0003850421144725844,
|
| 1303 |
+
"loss": 0.6559,
|
| 1304 |
+
"step": 18500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 126.96245733788396,
|
| 1308 |
+
"grad_norm": 0.5486589670181274,
|
| 1309 |
+
"learning_rate": 0.00038039645708488955,
|
| 1310 |
+
"loss": 0.6606,
|
| 1311 |
+
"step": 18600
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 127.64505119453925,
|
| 1315 |
+
"grad_norm": 0.44690069556236267,
|
| 1316 |
+
"learning_rate": 0.0003757717742777895,
|
| 1317 |
+
"loss": 0.6538,
|
| 1318 |
+
"step": 18700
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 128.32764505119454,
|
| 1322 |
+
"grad_norm": 0.5112203359603882,
|
| 1323 |
+
"learning_rate": 0.00037116863797791145,
|
| 1324 |
+
"loss": 0.6523,
|
| 1325 |
+
"step": 18800
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 129.01023890784984,
|
| 1329 |
+
"grad_norm": 0.479837030172348,
|
| 1330 |
+
"learning_rate": 0.0003665876174472628,
|
| 1331 |
+
"loss": 0.6563,
|
| 1332 |
+
"step": 18900
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 129.6928327645051,
|
| 1336 |
+
"grad_norm": 0.4698416292667389,
|
| 1337 |
+
"learning_rate": 0.00036202927921283144,
|
| 1338 |
+
"loss": 0.6495,
|
| 1339 |
+
"step": 19000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 130.37542662116041,
|
| 1343 |
+
"grad_norm": 0.5283064246177673,
|
| 1344 |
+
"learning_rate": 0.0003574941869965248,
|
| 1345 |
+
"loss": 0.6474,
|
| 1346 |
+
"step": 19100
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 131.0580204778157,
|
| 1350 |
+
"grad_norm": 0.4855287969112396,
|
| 1351 |
+
"learning_rate": 0.00035298290164545443,
|
| 1352 |
+
"loss": 0.652,
|
| 1353 |
+
"step": 19200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 131.740614334471,
|
| 1357 |
+
"grad_norm": 0.4482058882713318,
|
| 1358 |
+
"learning_rate": 0.00034849598106257795,
|
| 1359 |
+
"loss": 0.6454,
|
| 1360 |
+
"step": 19300
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 132.4232081911263,
|
| 1364 |
+
"grad_norm": 0.487212210893631,
|
| 1365 |
+
"learning_rate": 0.00034403398013770347,
|
| 1366 |
+
"loss": 0.643,
|
| 1367 |
+
"step": 19400
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 133.10580204778157,
|
| 1371 |
+
"grad_norm": 0.4655930697917938,
|
| 1372 |
+
"learning_rate": 0.0003395974506788674,
|
| 1373 |
+
"loss": 0.6477,
|
| 1374 |
+
"step": 19500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 133.78839590443687,
|
| 1378 |
+
"grad_norm": 0.4315454661846161,
|
| 1379 |
+
"learning_rate": 0.00033518694134409367,
|
| 1380 |
+
"loss": 0.6411,
|
| 1381 |
+
"step": 19600
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 134.47098976109214,
|
| 1385 |
+
"grad_norm": 0.505107581615448,
|
| 1386 |
+
"learning_rate": 0.00033080299757354216,
|
| 1387 |
+
"loss": 0.6396,
|
| 1388 |
+
"step": 19700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 135.15358361774744,
|
| 1392 |
+
"grad_norm": 0.47746554017066956,
|
| 1393 |
+
"learning_rate": 0.00032644616152205385,
|
| 1394 |
+
"loss": 0.6436,
|
| 1395 |
+
"step": 19800
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 135.83617747440272,
|
| 1399 |
+
"grad_norm": 0.44861286878585815,
|
| 1400 |
+
"learning_rate": 0.0003221169719921052,
|
| 1401 |
+
"loss": 0.6382,
|
| 1402 |
+
"step": 19900
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 136.51877133105802,
|
| 1406 |
+
"grad_norm": 0.4924197196960449,
|
| 1407 |
+
"learning_rate": 0.00031781596436717465,
|
| 1408 |
+
"loss": 0.636,
|
| 1409 |
+
"step": 20000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 137.20136518771332,
|
| 1413 |
+
"grad_norm": 0.4533005952835083,
|
| 1414 |
+
"learning_rate": 0.0003135436705455319,
|
| 1415 |
+
"loss": 0.6388,
|
| 1416 |
+
"step": 20100
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 137.8839590443686,
|
| 1420 |
+
"grad_norm": 0.4372631311416626,
|
| 1421 |
+
"learning_rate": 0.0003093006188744598,
|
| 1422 |
+
"loss": 0.6349,
|
| 1423 |
+
"step": 20200
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 138.5665529010239,
|
| 1427 |
+
"grad_norm": 0.5831769108772278,
|
| 1428 |
+
"learning_rate": 0.0003050873340849148,
|
| 1429 |
+
"loss": 0.6323,
|
| 1430 |
+
"step": 20300
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 139.24914675767917,
|
| 1434 |
+
"grad_norm": 0.41959720849990845,
|
| 1435 |
+
"learning_rate": 0.00030090433722663367,
|
| 1436 |
+
"loss": 0.6357,
|
| 1437 |
+
"step": 20400
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 139.93174061433447,
|
| 1441 |
+
"grad_norm": 0.47732993960380554,
|
| 1442 |
+
"learning_rate": 0.00029675214560369637,
|
| 1443 |
+
"loss": 0.6323,
|
| 1444 |
+
"step": 20500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 140.61433447098977,
|
| 1448 |
+
"grad_norm": 0.4794413447380066,
|
| 1449 |
+
"learning_rate": 0.0002926312727105518,
|
| 1450 |
+
"loss": 0.6303,
|
| 1451 |
+
"step": 20600
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 141.29692832764505,
|
| 1455 |
+
"grad_norm": 0.42916139960289,
|
| 1456 |
+
"learning_rate": 0.00028854222816851547,
|
| 1457 |
+
"loss": 0.6327,
|
| 1458 |
+
"step": 20700
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 141.97952218430035,
|
| 1462 |
+
"grad_norm": 0.42981767654418945,
|
| 1463 |
+
"learning_rate": 0.00028448551766274466,
|
| 1464 |
+
"loss": 0.6285,
|
| 1465 |
+
"step": 20800
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 142.66211604095562,
|
| 1469 |
+
"grad_norm": 0.48353445529937744,
|
| 1470 |
+
"learning_rate": 0.0002804616428797021,
|
| 1471 |
+
"loss": 0.6272,
|
| 1472 |
+
"step": 20900
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 143.34470989761093,
|
| 1476 |
+
"grad_norm": 0.4736007750034332,
|
| 1477 |
+
"learning_rate": 0.0002764711014451121,
|
| 1478 |
+
"loss": 0.6288,
|
| 1479 |
+
"step": 21000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 144.0273037542662,
|
| 1483 |
+
"grad_norm": 0.47045692801475525,
|
| 1484 |
+
"learning_rate": 0.0002725143868624209,
|
| 1485 |
+
"loss": 0.6257,
|
| 1486 |
+
"step": 21100
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 144.7098976109215,
|
| 1490 |
+
"grad_norm": 0.4963953495025635,
|
| 1491 |
+
"learning_rate": 0.0002685919884517659,
|
| 1492 |
+
"loss": 0.6251,
|
| 1493 |
+
"step": 21200
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 145.3924914675768,
|
| 1497 |
+
"grad_norm": 0.4388594925403595,
|
| 1498 |
+
"learning_rate": 0.00026470439128946214,
|
| 1499 |
+
"loss": 0.6249,
|
| 1500 |
+
"step": 21300
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 146.07508532423208,
|
| 1504 |
+
"grad_norm": 0.4757547378540039,
|
| 1505 |
+
"learning_rate": 0.0002608520761480135,
|
| 1506 |
+
"loss": 0.6241,
|
| 1507 |
+
"step": 21400
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 146.75767918088738,
|
| 1511 |
+
"grad_norm": 0.46713003516197205,
|
| 1512 |
+
"learning_rate": 0.00025703551943665636,
|
| 1513 |
+
"loss": 0.6229,
|
| 1514 |
+
"step": 21500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 147.44027303754265,
|
| 1518 |
+
"grad_norm": 0.43607574701309204,
|
| 1519 |
+
"learning_rate": 0.00025325519314244374,
|
| 1520 |
+
"loss": 0.623,
|
| 1521 |
+
"step": 21600
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 148.12286689419795,
|
| 1525 |
+
"grad_norm": 0.4766788184642792,
|
| 1526 |
+
"learning_rate": 0.00024951156477187413,
|
| 1527 |
+
"loss": 0.6212,
|
| 1528 |
+
"step": 21700
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 148.80546075085323,
|
| 1532 |
+
"grad_norm": 0.49822133779525757,
|
| 1533 |
+
"learning_rate": 0.00024580509729307673,
|
| 1534 |
+
"loss": 0.6215,
|
| 1535 |
+
"step": 21800
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 149.48805460750853,
|
| 1539 |
+
"grad_norm": 0.4069366455078125,
|
| 1540 |
+
"learning_rate": 0.0002421362490785569,
|
| 1541 |
+
"loss": 0.6199,
|
| 1542 |
+
"step": 21900
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 150.17064846416383,
|
| 1546 |
+
"grad_norm": 0.46552979946136475,
|
| 1547 |
+
"learning_rate": 0.00023850547384850974,
|
| 1548 |
+
"loss": 0.6194,
|
| 1549 |
+
"step": 22000
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 150.8532423208191,
|
| 1553 |
+
"grad_norm": 0.45079755783081055,
|
| 1554 |
+
"learning_rate": 0.00023491322061470915,
|
| 1555 |
+
"loss": 0.6202,
|
| 1556 |
+
"step": 22100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 151.5358361774744,
|
| 1560 |
+
"grad_norm": 0.46759602427482605,
|
| 1561 |
+
"learning_rate": 0.00023135993362497948,
|
| 1562 |
+
"loss": 0.6181,
|
| 1563 |
+
"step": 22200
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 152.21843003412968,
|
| 1567 |
+
"grad_norm": 0.45771974325180054,
|
| 1568 |
+
"learning_rate": 0.00022784605230825666,
|
| 1569 |
+
"loss": 0.6164,
|
| 1570 |
+
"step": 22300
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 152.90102389078498,
|
| 1574 |
+
"grad_norm": 0.5069631338119507,
|
| 1575 |
+
"learning_rate": 0.00022437201122024375,
|
| 1576 |
+
"loss": 0.6187,
|
| 1577 |
+
"step": 22400
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 153.5836177474403,
|
| 1581 |
+
"grad_norm": 0.43779122829437256,
|
| 1582 |
+
"learning_rate": 0.00022093823998967115,
|
| 1583 |
+
"loss": 0.6164,
|
| 1584 |
+
"step": 22500
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 154.26621160409556,
|
| 1588 |
+
"grad_norm": 0.477224200963974,
|
| 1589 |
+
"learning_rate": 0.00021754516326516458,
|
| 1590 |
+
"loss": 0.6151,
|
| 1591 |
+
"step": 22600
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 154.94880546075086,
|
| 1595 |
+
"grad_norm": 0.47925087809562683,
|
| 1596 |
+
"learning_rate": 0.00021419320066272973,
|
| 1597 |
+
"loss": 0.6166,
|
| 1598 |
+
"step": 22700
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 155.63139931740614,
|
| 1602 |
+
"grad_norm": 0.4455689787864685,
|
| 1603 |
+
"learning_rate": 0.00021088276671385882,
|
| 1604 |
+
"loss": 0.6139,
|
| 1605 |
+
"step": 22800
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 156.31399317406144,
|
| 1609 |
+
"grad_norm": 0.4499022364616394,
|
| 1610 |
+
"learning_rate": 0.00020761427081426677,
|
| 1611 |
+
"loss": 0.6132,
|
| 1612 |
+
"step": 22900
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 156.9965870307167,
|
| 1616 |
+
"grad_norm": 0.4519612193107605,
|
| 1617 |
+
"learning_rate": 0.00020438811717326112,
|
| 1618 |
+
"loss": 0.6154,
|
| 1619 |
+
"step": 23000
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 157.679180887372,
|
| 1623 |
+
"grad_norm": 0.4130764603614807,
|
| 1624 |
+
"learning_rate": 0.0002012047047637547,
|
| 1625 |
+
"loss": 0.6122,
|
| 1626 |
+
"step": 23100
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 158.36177474402731,
|
| 1630 |
+
"grad_norm": 0.44488415122032166,
|
| 1631 |
+
"learning_rate": 0.0001980644272729254,
|
| 1632 |
+
"loss": 0.6132,
|
| 1633 |
+
"step": 23200
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 159.0443686006826,
|
| 1637 |
+
"grad_norm": 0.47042757272720337,
|
| 1638 |
+
"learning_rate": 0.00019496767305352946,
|
| 1639 |
+
"loss": 0.6135,
|
| 1640 |
+
"step": 23300
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 159.7269624573379,
|
| 1644 |
+
"grad_norm": 0.3974725604057312,
|
| 1645 |
+
"learning_rate": 0.00019191482507587376,
|
| 1646 |
+
"loss": 0.6107,
|
| 1647 |
+
"step": 23400
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 160.40955631399316,
|
| 1651 |
+
"grad_norm": 0.46765318512916565,
|
| 1652 |
+
"learning_rate": 0.00018890626088045593,
|
| 1653 |
+
"loss": 0.6109,
|
| 1654 |
+
"step": 23500
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 161.09215017064847,
|
| 1658 |
+
"grad_norm": 0.42309921979904175,
|
| 1659 |
+
"learning_rate": 0.00018594235253127373,
|
| 1660 |
+
"loss": 0.6122,
|
| 1661 |
+
"step": 23600
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 161.77474402730377,
|
| 1665 |
+
"grad_norm": 0.3989265561103821,
|
| 1666 |
+
"learning_rate": 0.00018302346656981234,
|
| 1667 |
+
"loss": 0.6099,
|
| 1668 |
+
"step": 23700
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 162.45733788395904,
|
| 1672 |
+
"grad_norm": 0.4663792550563812,
|
| 1673 |
+
"learning_rate": 0.0001801499639697153,
|
| 1674 |
+
"loss": 0.6103,
|
| 1675 |
+
"step": 23800
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 163.13993174061434,
|
| 1679 |
+
"grad_norm": 0.3913436233997345,
|
| 1680 |
+
"learning_rate": 0.0001773222000921435,
|
| 1681 |
+
"loss": 0.6106,
|
| 1682 |
+
"step": 23900
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 163.82252559726962,
|
| 1686 |
+
"grad_norm": 0.41263365745544434,
|
| 1687 |
+
"learning_rate": 0.00017454052464182778,
|
| 1688 |
+
"loss": 0.6092,
|
| 1689 |
+
"step": 24000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 164.50511945392492,
|
| 1693 |
+
"grad_norm": 0.4633448123931885,
|
| 1694 |
+
"learning_rate": 0.0001718052816238222,
|
| 1695 |
+
"loss": 0.608,
|
| 1696 |
+
"step": 24100
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 165.1877133105802,
|
| 1700 |
+
"grad_norm": 0.3786383271217346,
|
| 1701 |
+
"learning_rate": 0.00016911680930096085,
|
| 1702 |
+
"loss": 0.6094,
|
| 1703 |
+
"step": 24200
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 165.8703071672355,
|
| 1707 |
+
"grad_norm": 0.4065646529197693,
|
| 1708 |
+
"learning_rate": 0.00016647544015202623,
|
| 1709 |
+
"loss": 0.608,
|
| 1710 |
+
"step": 24300
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 166.5529010238908,
|
| 1714 |
+
"grad_norm": 0.43729546666145325,
|
| 1715 |
+
"learning_rate": 0.00016388150083063117,
|
| 1716 |
+
"loss": 0.6072,
|
| 1717 |
+
"step": 24400
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 167.23549488054607,
|
| 1721 |
+
"grad_norm": 0.4111308157444,
|
| 1722 |
+
"learning_rate": 0.00016133531212482325,
|
| 1723 |
+
"loss": 0.6078,
|
| 1724 |
+
"step": 24500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 167.91808873720137,
|
| 1728 |
+
"grad_norm": 0.39553767442703247,
|
| 1729 |
+
"learning_rate": 0.00015883718891741258,
|
| 1730 |
+
"loss": 0.6066,
|
| 1731 |
+
"step": 24600
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 168.60068259385665,
|
| 1735 |
+
"grad_norm": 0.4074500501155853,
|
| 1736 |
+
"learning_rate": 0.00015638744014703128,
|
| 1737 |
+
"loss": 0.6068,
|
| 1738 |
+
"step": 24700
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 169.28327645051195,
|
| 1742 |
+
"grad_norm": 0.42327943444252014,
|
| 1743 |
+
"learning_rate": 0.0001539863687699274,
|
| 1744 |
+
"loss": 0.607,
|
| 1745 |
+
"step": 24800
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 169.96587030716722,
|
| 1749 |
+
"grad_norm": 0.40863099694252014,
|
| 1750 |
+
"learning_rate": 0.0001516342717224992,
|
| 1751 |
+
"loss": 0.6067,
|
| 1752 |
+
"step": 24900
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 170.64846416382252,
|
| 1756 |
+
"grad_norm": 0.4853709638118744,
|
| 1757 |
+
"learning_rate": 0.00014933143988457237,
|
| 1758 |
+
"loss": 0.6056,
|
| 1759 |
+
"step": 25000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 171.33105802047783,
|
| 1763 |
+
"grad_norm": 0.383801132440567,
|
| 1764 |
+
"learning_rate": 0.00014707815804342907,
|
| 1765 |
+
"loss": 0.6063,
|
| 1766 |
+
"step": 25100
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 172.0136518771331,
|
| 1770 |
+
"grad_norm": 0.4241873621940613,
|
| 1771 |
+
"learning_rate": 0.00014487470485858767,
|
| 1772 |
+
"loss": 0.6058,
|
| 1773 |
+
"step": 25200
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 172.6962457337884,
|
| 1777 |
+
"grad_norm": 0.4262351989746094,
|
| 1778 |
+
"learning_rate": 0.00014272135282734167,
|
| 1779 |
+
"loss": 0.6043,
|
| 1780 |
+
"step": 25300
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 173.37883959044368,
|
| 1784 |
+
"grad_norm": 0.38850197196006775,
|
| 1785 |
+
"learning_rate": 0.00014061836825106022,
|
| 1786 |
+
"loss": 0.606,
|
| 1787 |
+
"step": 25400
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 174.06143344709898,
|
| 1791 |
+
"grad_norm": 0.4169691801071167,
|
| 1792 |
+
"learning_rate": 0.00013856601120225574,
|
| 1793 |
+
"loss": 0.6038,
|
| 1794 |
+
"step": 25500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 174.74402730375428,
|
| 1798 |
+
"grad_norm": 0.43140584230422974,
|
| 1799 |
+
"learning_rate": 0.00013656453549242055,
|
| 1800 |
+
"loss": 0.6043,
|
| 1801 |
+
"step": 25600
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 175.42662116040955,
|
| 1805 |
+
"grad_norm": 0.3729684054851532,
|
| 1806 |
+
"learning_rate": 0.00013461418864063861,
|
| 1807 |
+
"loss": 0.6047,
|
| 1808 |
+
"step": 25700
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 176.10921501706486,
|
| 1812 |
+
"grad_norm": 0.435611367225647,
|
| 1813 |
+
"learning_rate": 0.00013271521184297523,
|
| 1814 |
+
"loss": 0.6035,
|
| 1815 |
+
"step": 25800
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 176.79180887372013,
|
| 1819 |
+
"grad_norm": 0.4359755218029022,
|
| 1820 |
+
"learning_rate": 0.00013086783994264905,
|
| 1821 |
+
"loss": 0.6044,
|
| 1822 |
+
"step": 25900
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 177.47440273037543,
|
| 1826 |
+
"grad_norm": 0.3867483139038086,
|
| 1827 |
+
"learning_rate": 0.00012907230140098895,
|
| 1828 |
+
"loss": 0.6031,
|
| 1829 |
+
"step": 26000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 178.1569965870307,
|
| 1833 |
+
"grad_norm": 0.41799867153167725,
|
| 1834 |
+
"learning_rate": 0.00012732881826918094,
|
| 1835 |
+
"loss": 0.6031,
|
| 1836 |
+
"step": 26100
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 178.839590443686,
|
| 1840 |
+
"grad_norm": 0.43031203746795654,
|
| 1841 |
+
"learning_rate": 0.00012563760616080745,
|
| 1842 |
+
"loss": 0.6028,
|
| 1843 |
+
"step": 26200
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 179.5221843003413,
|
| 1847 |
+
"grad_norm": 0.3603518605232239,
|
| 1848 |
+
"learning_rate": 0.00012399887422518243,
|
| 1849 |
+
"loss": 0.6026,
|
| 1850 |
+
"step": 26300
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 180.20477815699658,
|
| 1854 |
+
"grad_norm": 0.42124807834625244,
|
| 1855 |
+
"learning_rate": 0.0001224128251214867,
|
| 1856 |
+
"loss": 0.6021,
|
| 1857 |
+
"step": 26400
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 180.88737201365188,
|
| 1861 |
+
"grad_norm": 0.40946000814437866,
|
| 1862 |
+
"learning_rate": 0.00012087965499370513,
|
| 1863 |
+
"loss": 0.6036,
|
| 1864 |
+
"step": 26500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 181.56996587030716,
|
| 1868 |
+
"grad_norm": 0.3763185143470764,
|
| 1869 |
+
"learning_rate": 0.00011939955344636983,
|
| 1870 |
+
"loss": 0.6022,
|
| 1871 |
+
"step": 26600
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 182.25255972696246,
|
| 1875 |
+
"grad_norm": 0.4167167544364929,
|
| 1876 |
+
"learning_rate": 0.00011797270352111224,
|
| 1877 |
+
"loss": 0.6016,
|
| 1878 |
+
"step": 26700
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 182.93515358361773,
|
| 1882 |
+
"grad_norm": 0.42429491877555847,
|
| 1883 |
+
"learning_rate": 0.0001165992816740268,
|
| 1884 |
+
"loss": 0.6021,
|
| 1885 |
+
"step": 26800
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 183.61774744027304,
|
| 1889 |
+
"grad_norm": 0.38188785314559937,
|
| 1890 |
+
"learning_rate": 0.00011527945775384831,
|
| 1891 |
+
"loss": 0.6022,
|
| 1892 |
+
"step": 26900
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 184.30034129692834,
|
| 1896 |
+
"grad_norm": 0.47042107582092285,
|
| 1897 |
+
"learning_rate": 0.00011401339498094774,
|
| 1898 |
+
"loss": 0.6016,
|
| 1899 |
+
"step": 27000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 184.9829351535836,
|
| 1903 |
+
"grad_norm": 0.422690212726593,
|
| 1904 |
+
"learning_rate": 0.00011280124992714677,
|
| 1905 |
+
"loss": 0.602,
|
| 1906 |
+
"step": 27100
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 185.6655290102389,
|
| 1910 |
+
"grad_norm": 0.3714023232460022,
|
| 1911 |
+
"learning_rate": 0.00011164317249635475,
|
| 1912 |
+
"loss": 0.6014,
|
| 1913 |
+
"step": 27200
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 186.3481228668942,
|
| 1917 |
+
"grad_norm": 0.43933653831481934,
|
| 1918 |
+
"learning_rate": 0.00011053930590603041,
|
| 1919 |
+
"loss": 0.6003,
|
| 1920 |
+
"step": 27300
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 187.0307167235495,
|
| 1924 |
+
"grad_norm": 0.3583015501499176,
|
| 1925 |
+
"learning_rate": 0.00010948978666947039,
|
| 1926 |
+
"loss": 0.6016,
|
| 1927 |
+
"step": 27400
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 187.7133105802048,
|
| 1931 |
+
"grad_norm": 0.38929620385169983,
|
| 1932 |
+
"learning_rate": 0.00010849474457892682,
|
| 1933 |
+
"loss": 0.6006,
|
| 1934 |
+
"step": 27500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 188.39590443686006,
|
| 1938 |
+
"grad_norm": 0.42569249868392944,
|
| 1939 |
+
"learning_rate": 0.0001075543026895563,
|
| 1940 |
+
"loss": 0.6008,
|
| 1941 |
+
"step": 27600
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 189.07849829351537,
|
| 1945 |
+
"grad_norm": 0.37430456280708313,
|
| 1946 |
+
"learning_rate": 0.00010666857730420167,
|
| 1947 |
+
"loss": 0.601,
|
| 1948 |
+
"step": 27700
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 189.76109215017064,
|
| 1952 |
+
"grad_norm": 0.4152965545654297,
|
| 1953 |
+
"learning_rate": 0.00010583767795900921,
|
| 1954 |
+
"loss": 0.6006,
|
| 1955 |
+
"step": 27800
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 190.44368600682594,
|
| 1959 |
+
"grad_norm": 0.4082518517971039,
|
| 1960 |
+
"learning_rate": 0.00010506170740988232,
|
| 1961 |
+
"loss": 0.5998,
|
| 1962 |
+
"step": 27900
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 191.12627986348122,
|
| 1966 |
+
"grad_norm": 0.3863820731639862,
|
| 1967 |
+
"learning_rate": 0.00010434076161977406,
|
| 1968 |
+
"loss": 0.6013,
|
| 1969 |
+
"step": 28000
|
| 1970 |
+
}
|
| 1971 |
+
],
|
| 1972 |
+
"logging_steps": 100,
|
| 1973 |
+
"max_steps": 29200,
|
| 1974 |
+
"num_input_tokens_seen": 0,
|
| 1975 |
+
"num_train_epochs": 200,
|
| 1976 |
+
"save_steps": 500,
|
| 1977 |
+
"stateful_callbacks": {
|
| 1978 |
+
"TrainerControl": {
|
| 1979 |
+
"args": {
|
| 1980 |
+
"should_epoch_stop": false,
|
| 1981 |
+
"should_evaluate": false,
|
| 1982 |
+
"should_log": false,
|
| 1983 |
+
"should_save": true,
|
| 1984 |
+
"should_training_stop": false
|
| 1985 |
+
},
|
| 1986 |
+
"attributes": {}
|
| 1987 |
+
}
|
| 1988 |
+
},
|
| 1989 |
+
"total_flos": 8.957241596387e+17,
|
| 1990 |
+
"train_batch_size": 256,
|
| 1991 |
+
"trial_name": null,
|
| 1992 |
+
"trial_params": null
|
| 1993 |
+
}
|
checkpoint-28000/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b3988ed6f3d05a9a0d746b5535dfdb8d9bd5cdf3d3059847445b0b1d44dcf65
|
| 3 |
+
size 5240
|
checkpoint-28000/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-28500/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_function": "gelu_new",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPTJForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attn_pdrop": 0.1,
|
| 7 |
+
"bos_token_id": 50256,
|
| 8 |
+
"embd_pdrop": 0.1,
|
| 9 |
+
"eos_token_id": 50256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"layer_norm_epsilon": 1e-05,
|
| 12 |
+
"model_type": "gptj",
|
| 13 |
+
"n_embd": 768,
|
| 14 |
+
"n_head": 12,
|
| 15 |
+
"n_inner": null,
|
| 16 |
+
"n_layer": 12,
|
| 17 |
+
"n_positions": 1024,
|
| 18 |
+
"resid_pdrop": 0.1,
|
| 19 |
+
"rotary_dim": 64,
|
| 20 |
+
"tie_word_embeddings": false,
|
| 21 |
+
"torch_dtype": "float32",
|
| 22 |
+
"transformers_version": "4.46.2",
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 50257
|
| 25 |
+
}
|
checkpoint-28500/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.46.2"
|
| 6 |
+
}
|
checkpoint-28500/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-28500/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6568728ec17962f4daada53b40ae2490ca1ee51cbdcf6607eb76af6066afe3c4
|
| 3 |
+
size 648995876
|
checkpoint-28500/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b6bac70500f4f2772ddb4338cfd1bb324e80308af81d0f81e4f2d498faaec326
|
| 3 |
+
size 1298039738
|
checkpoint-28500/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:425d4c80a123f104a129d064df2054f04c525f97d6eb5ed07b3b8e2944bfdace
|
| 3 |
+
size 14512
|
checkpoint-28500/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:43186e5b8a70005b8e9d542a95d0a638cdeee627579c0021f40d41e139e0da42
|
| 3 |
+
size 14512
|
checkpoint-28500/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1edec7be15fcf0ca3447e617b0f3ef6d0fc6940f4e4f9c4e0278dfc5f3194519
|
| 3 |
+
size 1000
|
checkpoint-28500/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|endoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endoftext|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "<|endoftext|>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<|endoftext|>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": true,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-28500/tokenizer_config.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"50256": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": true,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"bos_token": "<|endoftext|>",
|
| 15 |
+
"clean_up_tokenization_spaces": false,
|
| 16 |
+
"eos_token": "<|endoftext|>",
|
| 17 |
+
"errors": "replace",
|
| 18 |
+
"model_max_length": 1024,
|
| 19 |
+
"pad_token": "<|endoftext|>",
|
| 20 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 21 |
+
"unk_token": "<|endoftext|>"
|
| 22 |
+
}
|
checkpoint-28500/trainer_state.json
ADDED
|
@@ -0,0 +1,2028 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 194.53924914675767,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 28500,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.6825938566552902,
|
| 13 |
+
"grad_norm": 1.5864567756652832,
|
| 14 |
+
"learning_rate": 0.0001,
|
| 15 |
+
"loss": 5.292,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.36518771331058,
|
| 20 |
+
"grad_norm": 0.6435245275497437,
|
| 21 |
+
"learning_rate": 0.0002,
|
| 22 |
+
"loss": 1.3885,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 2.04778156996587,
|
| 27 |
+
"grad_norm": 0.6205620169639587,
|
| 28 |
+
"learning_rate": 0.0003,
|
| 29 |
+
"loss": 1.2326,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 2.73037542662116,
|
| 34 |
+
"grad_norm": 0.668118953704834,
|
| 35 |
+
"learning_rate": 0.0004,
|
| 36 |
+
"loss": 1.234,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 3.4129692832764507,
|
| 41 |
+
"grad_norm": 0.5799328684806824,
|
| 42 |
+
"learning_rate": 0.0005,
|
| 43 |
+
"loss": 1.2338,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 4.09556313993174,
|
| 48 |
+
"grad_norm": 0.5505828261375427,
|
| 49 |
+
"learning_rate": 0.0006,
|
| 50 |
+
"loss": 1.2287,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 4.778156996587031,
|
| 55 |
+
"grad_norm": 0.515738844871521,
|
| 56 |
+
"learning_rate": 0.0007,
|
| 57 |
+
"loss": 1.2327,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 5.460750853242321,
|
| 62 |
+
"grad_norm": 0.509695291519165,
|
| 63 |
+
"learning_rate": 0.0008,
|
| 64 |
+
"loss": 1.2565,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 6.143344709897611,
|
| 69 |
+
"grad_norm": 0.6063594818115234,
|
| 70 |
+
"learning_rate": 0.0009000000000000001,
|
| 71 |
+
"loss": 1.2289,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 6.825938566552901,
|
| 76 |
+
"grad_norm": 0.5587486624717712,
|
| 77 |
+
"learning_rate": 0.001,
|
| 78 |
+
"loss": 1.2334,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 7.508532423208191,
|
| 83 |
+
"grad_norm": 0.5273250341415405,
|
| 84 |
+
"learning_rate": 0.0009999721746341338,
|
| 85 |
+
"loss": 1.2316,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 8.19112627986348,
|
| 90 |
+
"grad_norm": 0.593492329120636,
|
| 91 |
+
"learning_rate": 0.000999888701977651,
|
| 92 |
+
"loss": 1.2279,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 8.873720136518772,
|
| 97 |
+
"grad_norm": 0.44249382615089417,
|
| 98 |
+
"learning_rate": 0.000999749592353472,
|
| 99 |
+
"loss": 1.231,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 9.556313993174061,
|
| 104 |
+
"grad_norm": 0.4120616018772125,
|
| 105 |
+
"learning_rate": 0.0009995548629650466,
|
| 106 |
+
"loss": 1.2282,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 10.238907849829351,
|
| 111 |
+
"grad_norm": 0.5016219019889832,
|
| 112 |
+
"learning_rate": 0.0009993045378942262,
|
| 113 |
+
"loss": 1.2261,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 10.921501706484642,
|
| 118 |
+
"grad_norm": 0.5626952648162842,
|
| 119 |
+
"learning_rate": 0.0009989986480982847,
|
| 120 |
+
"loss": 1.2294,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 11.604095563139932,
|
| 125 |
+
"grad_norm": 0.45088785886764526,
|
| 126 |
+
"learning_rate": 0.000998637231406091,
|
| 127 |
+
"loss": 1.2268,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 12.286689419795222,
|
| 132 |
+
"grad_norm": 0.39686086773872375,
|
| 133 |
+
"learning_rate": 0.0009982203325134304,
|
| 134 |
+
"loss": 1.2254,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 12.969283276450511,
|
| 139 |
+
"grad_norm": 0.3904390037059784,
|
| 140 |
+
"learning_rate": 0.0009977480029774771,
|
| 141 |
+
"loss": 1.2274,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 13.651877133105803,
|
| 146 |
+
"grad_norm": 0.38391488790512085,
|
| 147 |
+
"learning_rate": 0.0009972203012104186,
|
| 148 |
+
"loss": 1.2247,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 14.334470989761092,
|
| 153 |
+
"grad_norm": 0.4345687925815582,
|
| 154 |
+
"learning_rate": 0.0009966372924722312,
|
| 155 |
+
"loss": 1.2252,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 15.017064846416382,
|
| 160 |
+
"grad_norm": 0.34943118691444397,
|
| 161 |
+
"learning_rate": 0.0009959990488626104,
|
| 162 |
+
"loss": 1.2266,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 15.699658703071673,
|
| 167 |
+
"grad_norm": 0.3420363962650299,
|
| 168 |
+
"learning_rate": 0.0009953056493120537,
|
| 169 |
+
"loss": 1.224,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 16.38225255972696,
|
| 174 |
+
"grad_norm": 0.3876005709171295,
|
| 175 |
+
"learning_rate": 0.0009945571795721,
|
| 176 |
+
"loss": 1.2241,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 17.064846416382252,
|
| 181 |
+
"grad_norm": 0.3726520836353302,
|
| 182 |
+
"learning_rate": 0.0009937537322047237,
|
| 183 |
+
"loss": 1.2258,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 17.747440273037544,
|
| 188 |
+
"grad_norm": 0.4071378707885742,
|
| 189 |
+
"learning_rate": 0.0009928954065708895,
|
| 190 |
+
"loss": 1.2235,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 18.43003412969283,
|
| 195 |
+
"grad_norm": 0.443796843290329,
|
| 196 |
+
"learning_rate": 0.0009919823088182628,
|
| 197 |
+
"loss": 1.2346,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 19.112627986348123,
|
| 202 |
+
"grad_norm": 0.3463057279586792,
|
| 203 |
+
"learning_rate": 0.000991014551868084,
|
| 204 |
+
"loss": 1.2292,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 19.795221843003414,
|
| 209 |
+
"grad_norm": 0.31187331676483154,
|
| 210 |
+
"learning_rate": 0.0009899922554012027,
|
| 211 |
+
"loss": 1.2237,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 20.477815699658702,
|
| 216 |
+
"grad_norm": 0.3052064776420593,
|
| 217 |
+
"learning_rate": 0.0009889155458432776,
|
| 218 |
+
"loss": 1.2242,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 21.160409556313994,
|
| 223 |
+
"grad_norm": 0.3126174807548523,
|
| 224 |
+
"learning_rate": 0.000987784556349142,
|
| 225 |
+
"loss": 1.2251,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 21.843003412969285,
|
| 230 |
+
"grad_norm": 0.29788631200790405,
|
| 231 |
+
"learning_rate": 0.0009865994267863351,
|
| 232 |
+
"loss": 1.2224,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 22.525597269624573,
|
| 237 |
+
"grad_norm": 0.2966502606868744,
|
| 238 |
+
"learning_rate": 0.000985360303717807,
|
| 239 |
+
"loss": 1.2233,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 23.208191126279864,
|
| 244 |
+
"grad_norm": 0.31454625725746155,
|
| 245 |
+
"learning_rate": 0.0009840673403837918,
|
| 246 |
+
"loss": 1.2242,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 23.890784982935152,
|
| 251 |
+
"grad_norm": 0.33144181966781616,
|
| 252 |
+
"learning_rate": 0.0009827206966828582,
|
| 253 |
+
"loss": 1.2219,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 24.573378839590443,
|
| 258 |
+
"grad_norm": 0.3603832423686981,
|
| 259 |
+
"learning_rate": 0.000981320539152133,
|
| 260 |
+
"loss": 1.2227,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 25.255972696245735,
|
| 265 |
+
"grad_norm": 0.28107529878616333,
|
| 266 |
+
"learning_rate": 0.0009798670409467075,
|
| 267 |
+
"loss": 1.2236,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 25.938566552901023,
|
| 272 |
+
"grad_norm": 0.3241814076900482,
|
| 273 |
+
"learning_rate": 0.0009783603818182237,
|
| 274 |
+
"loss": 1.2212,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 26.621160409556314,
|
| 279 |
+
"grad_norm": 0.35018429160118103,
|
| 280 |
+
"learning_rate": 0.0009768007480926437,
|
| 281 |
+
"loss": 1.2223,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 27.303754266211605,
|
| 286 |
+
"grad_norm": 0.2918601632118225,
|
| 287 |
+
"learning_rate": 0.0009751883326472076,
|
| 288 |
+
"loss": 1.2228,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 27.986348122866893,
|
| 293 |
+
"grad_norm": 0.29299041628837585,
|
| 294 |
+
"learning_rate": 0.0009735233348865806,
|
| 295 |
+
"loss": 1.2208,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 28.668941979522184,
|
| 300 |
+
"grad_norm": 0.3291133642196655,
|
| 301 |
+
"learning_rate": 0.0009718059607181931,
|
| 302 |
+
"loss": 1.2219,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 29.351535836177476,
|
| 307 |
+
"grad_norm": 0.2924804985523224,
|
| 308 |
+
"learning_rate": 0.0009700364225267769,
|
| 309 |
+
"loss": 1.2222,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 30.034129692832764,
|
| 314 |
+
"grad_norm": 0.3081340193748474,
|
| 315 |
+
"learning_rate": 0.0009682149391480986,
|
| 316 |
+
"loss": 1.2201,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 30.716723549488055,
|
| 321 |
+
"grad_norm": 0.2960987091064453,
|
| 322 |
+
"learning_rate": 0.0009663417358418981,
|
| 323 |
+
"loss": 1.2213,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 31.399317406143346,
|
| 328 |
+
"grad_norm": 0.29421886801719666,
|
| 329 |
+
"learning_rate": 0.0009644170442640303,
|
| 330 |
+
"loss": 1.2201,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 32.081911262798634,
|
| 335 |
+
"grad_norm": 0.3381507098674774,
|
| 336 |
+
"learning_rate": 0.0009624411024378166,
|
| 337 |
+
"loss": 1.2186,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 32.76450511945392,
|
| 342 |
+
"grad_norm": 0.3033551275730133,
|
| 343 |
+
"learning_rate": 0.0009604141547246092,
|
| 344 |
+
"loss": 1.2199,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 33.44709897610922,
|
| 349 |
+
"grad_norm": 0.3156335949897766,
|
| 350 |
+
"learning_rate": 0.0009583364517935713,
|
| 351 |
+
"loss": 1.2189,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 34.129692832764505,
|
| 356 |
+
"grad_norm": 0.33952537178993225,
|
| 357 |
+
"learning_rate": 0.0009562082505906771,
|
| 358 |
+
"loss": 1.2163,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 34.81228668941979,
|
| 363 |
+
"grad_norm": 0.2941412925720215,
|
| 364 |
+
"learning_rate": 0.0009540298143069361,
|
| 365 |
+
"loss": 1.2179,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 35.49488054607509,
|
| 370 |
+
"grad_norm": 0.28668472170829773,
|
| 371 |
+
"learning_rate": 0.0009518014123458442,
|
| 372 |
+
"loss": 1.2158,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 36.177474402730375,
|
| 377 |
+
"grad_norm": 0.2789587080478668,
|
| 378 |
+
"learning_rate": 0.0009495233202900674,
|
| 379 |
+
"loss": 1.213,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 36.86006825938566,
|
| 384 |
+
"grad_norm": 0.3137347102165222,
|
| 385 |
+
"learning_rate": 0.0009471958198673614,
|
| 386 |
+
"loss": 1.214,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 37.54266211604096,
|
| 391 |
+
"grad_norm": 0.25685933232307434,
|
| 392 |
+
"learning_rate": 0.0009448191989157294,
|
| 393 |
+
"loss": 1.21,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 38.225255972696246,
|
| 398 |
+
"grad_norm": 0.3252381980419159,
|
| 399 |
+
"learning_rate": 0.0009423937513478271,
|
| 400 |
+
"loss": 1.2071,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 38.907849829351534,
|
| 405 |
+
"grad_norm": 0.2932053208351135,
|
| 406 |
+
"learning_rate": 0.0009399197771146144,
|
| 407 |
+
"loss": 1.2077,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 39.59044368600683,
|
| 412 |
+
"grad_norm": 0.33668607473373413,
|
| 413 |
+
"learning_rate": 0.0009373975821682607,
|
| 414 |
+
"loss": 1.2019,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 40.27303754266212,
|
| 419 |
+
"grad_norm": 0.31952792406082153,
|
| 420 |
+
"learning_rate": 0.0009348274784243094,
|
| 421 |
+
"loss": 1.1979,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 40.955631399317404,
|
| 426 |
+
"grad_norm": 0.3525752127170563,
|
| 427 |
+
"learning_rate": 0.0009322097837231025,
|
| 428 |
+
"loss": 1.1999,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 41.6382252559727,
|
| 433 |
+
"grad_norm": 0.3389936685562134,
|
| 434 |
+
"learning_rate": 0.0009295448217904745,
|
| 435 |
+
"loss": 1.1923,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 42.32081911262799,
|
| 440 |
+
"grad_norm": 0.3114762008190155,
|
| 441 |
+
"learning_rate": 0.0009268329221977179,
|
| 442 |
+
"loss": 1.19,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 43.003412969283275,
|
| 447 |
+
"grad_norm": 0.3166259229183197,
|
| 448 |
+
"learning_rate": 0.0009240744203208256,
|
| 449 |
+
"loss": 1.1914,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 43.68600682593857,
|
| 454 |
+
"grad_norm": 0.31299668550491333,
|
| 455 |
+
"learning_rate": 0.0009212696572990148,
|
| 456 |
+
"loss": 1.1843,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 44.36860068259386,
|
| 461 |
+
"grad_norm": 0.36728623509407043,
|
| 462 |
+
"learning_rate": 0.0009184189799925403,
|
| 463 |
+
"loss": 1.1828,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 45.051194539249146,
|
| 468 |
+
"grad_norm": 0.25386592745780945,
|
| 469 |
+
"learning_rate": 0.0009155227409397977,
|
| 470 |
+
"loss": 1.1849,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 45.73378839590444,
|
| 475 |
+
"grad_norm": 0.3170667290687561,
|
| 476 |
+
"learning_rate": 0.0009125812983137259,
|
| 477 |
+
"loss": 1.1789,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 46.41638225255973,
|
| 482 |
+
"grad_norm": 0.3354248106479645,
|
| 483 |
+
"learning_rate": 0.0009095950158775128,
|
| 484 |
+
"loss": 1.1781,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 47.098976109215016,
|
| 489 |
+
"grad_norm": 0.2761414647102356,
|
| 490 |
+
"learning_rate": 0.0009065642629396087,
|
| 491 |
+
"loss": 1.1803,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 47.781569965870304,
|
| 496 |
+
"grad_norm": 0.28061142563819885,
|
| 497 |
+
"learning_rate": 0.0009034894143080554,
|
| 498 |
+
"loss": 1.175,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 48.4641638225256,
|
| 503 |
+
"grad_norm": 0.3548315763473511,
|
| 504 |
+
"learning_rate": 0.0009003708502441333,
|
| 505 |
+
"loss": 1.1743,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 49.14675767918089,
|
| 510 |
+
"grad_norm": 0.3077629804611206,
|
| 511 |
+
"learning_rate": 0.0008972089564153364,
|
| 512 |
+
"loss": 1.1762,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 49.829351535836174,
|
| 517 |
+
"grad_norm": 0.3600843846797943,
|
| 518 |
+
"learning_rate": 0.0008940041238476759,
|
| 519 |
+
"loss": 1.1718,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 50.51194539249147,
|
| 524 |
+
"grad_norm": 0.3295997083187103,
|
| 525 |
+
"learning_rate": 0.0008907567488773241,
|
| 526 |
+
"loss": 1.1709,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 51.19453924914676,
|
| 531 |
+
"grad_norm": 0.32083791494369507,
|
| 532 |
+
"learning_rate": 0.0008874672331015996,
|
| 533 |
+
"loss": 1.1727,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 51.877133105802045,
|
| 538 |
+
"grad_norm": 0.32850903272628784,
|
| 539 |
+
"learning_rate": 0.0008841359833293022,
|
| 540 |
+
"loss": 1.1685,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 52.55972696245734,
|
| 545 |
+
"grad_norm": 0.33856746554374695,
|
| 546 |
+
"learning_rate": 0.000880763411530404,
|
| 547 |
+
"loss": 1.168,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 53.24232081911263,
|
| 552 |
+
"grad_norm": 0.34316346049308777,
|
| 553 |
+
"learning_rate": 0.0008773499347851014,
|
| 554 |
+
"loss": 1.1687,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 53.924914675767916,
|
| 559 |
+
"grad_norm": 0.38703614473342896,
|
| 560 |
+
"learning_rate": 0.0008738959752322358,
|
| 561 |
+
"loss": 1.165,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 54.60750853242321,
|
| 566 |
+
"grad_norm": 0.3954058587551117,
|
| 567 |
+
"learning_rate": 0.0008704019600170884,
|
| 568 |
+
"loss": 1.1637,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 55.2901023890785,
|
| 573 |
+
"grad_norm": 0.33417293429374695,
|
| 574 |
+
"learning_rate": 0.0008668683212385556,
|
| 575 |
+
"loss": 1.1635,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 55.972696245733786,
|
| 580 |
+
"grad_norm": 0.3305743932723999,
|
| 581 |
+
"learning_rate": 0.0008632954958957123,
|
| 582 |
+
"loss": 1.1601,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 56.65529010238908,
|
| 587 |
+
"grad_norm": 0.3755279779434204,
|
| 588 |
+
"learning_rate": 0.0008596839258337689,
|
| 589 |
+
"loss": 1.158,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 57.33788395904437,
|
| 594 |
+
"grad_norm": 0.3553084433078766,
|
| 595 |
+
"learning_rate": 0.000856034057689429,
|
| 596 |
+
"loss": 1.1568,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 58.02047781569966,
|
| 601 |
+
"grad_norm": 0.38361021876335144,
|
| 602 |
+
"learning_rate": 0.0008523463428356549,
|
| 603 |
+
"loss": 1.154,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 58.70307167235495,
|
| 608 |
+
"grad_norm": 0.35759079456329346,
|
| 609 |
+
"learning_rate": 0.0008486212373258465,
|
| 610 |
+
"loss": 1.152,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 59.38566552901024,
|
| 615 |
+
"grad_norm": 0.32452645897865295,
|
| 616 |
+
"learning_rate": 0.0008448592018374424,
|
| 617 |
+
"loss": 1.1497,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 60.06825938566553,
|
| 622 |
+
"grad_norm": 0.37652239203453064,
|
| 623 |
+
"learning_rate": 0.0008410607016149491,
|
| 624 |
+
"loss": 1.1461,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 60.75085324232082,
|
| 629 |
+
"grad_norm": 0.35698965191841125,
|
| 630 |
+
"learning_rate": 0.0008372262064124035,
|
| 631 |
+
"loss": 1.1453,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 61.43344709897611,
|
| 636 |
+
"grad_norm": 0.3505449891090393,
|
| 637 |
+
"learning_rate": 0.0008333561904352813,
|
| 638 |
+
"loss": 1.1409,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 62.1160409556314,
|
| 643 |
+
"grad_norm": 0.3646785616874695,
|
| 644 |
+
"learning_rate": 0.000829451132281851,
|
| 645 |
+
"loss": 1.1363,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 62.79863481228669,
|
| 650 |
+
"grad_norm": 0.34945082664489746,
|
| 651 |
+
"learning_rate": 0.0008255115148839877,
|
| 652 |
+
"loss": 1.1371,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 63.48122866894198,
|
| 657 |
+
"grad_norm": 0.34238216280937195,
|
| 658 |
+
"learning_rate": 0.0008215378254474485,
|
| 659 |
+
"loss": 1.1296,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 64.16382252559727,
|
| 664 |
+
"grad_norm": 0.37401944398880005,
|
| 665 |
+
"learning_rate": 0.0008175305553916226,
|
| 666 |
+
"loss": 1.1241,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 64.84641638225256,
|
| 671 |
+
"grad_norm": 0.3763269782066345,
|
| 672 |
+
"learning_rate": 0.000813490200288756,
|
| 673 |
+
"loss": 1.1257,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 65.52901023890784,
|
| 678 |
+
"grad_norm": 0.3553602993488312,
|
| 679 |
+
"learning_rate": 0.0008094172598026662,
|
| 680 |
+
"loss": 1.1147,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 66.21160409556315,
|
| 685 |
+
"grad_norm": 0.36700302362442017,
|
| 686 |
+
"learning_rate": 0.0008053122376269497,
|
| 687 |
+
"loss": 1.1078,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 66.89419795221843,
|
| 692 |
+
"grad_norm": 0.4266979694366455,
|
| 693 |
+
"learning_rate": 0.0008011756414226908,
|
| 694 |
+
"loss": 1.1114,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 67.57679180887372,
|
| 699 |
+
"grad_norm": 0.3603265583515167,
|
| 700 |
+
"learning_rate": 0.0007970079827556795,
|
| 701 |
+
"loss": 1.0967,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 68.25938566552901,
|
| 706 |
+
"grad_norm": 0.39329585433006287,
|
| 707 |
+
"learning_rate": 0.000792809777033148,
|
| 708 |
+
"loss": 1.0898,
|
| 709 |
+
"step": 10000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 68.9419795221843,
|
| 713 |
+
"grad_norm": 0.40609198808670044,
|
| 714 |
+
"learning_rate": 0.0007885815434400306,
|
| 715 |
+
"loss": 1.0944,
|
| 716 |
+
"step": 10100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 69.62457337883959,
|
| 720 |
+
"grad_norm": 0.3622093200683594,
|
| 721 |
+
"learning_rate": 0.0007843238048747572,
|
| 722 |
+
"loss": 1.0761,
|
| 723 |
+
"step": 10200
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 70.30716723549489,
|
| 727 |
+
"grad_norm": 0.39394432306289673,
|
| 728 |
+
"learning_rate": 0.0007800370878845866,
|
| 729 |
+
"loss": 1.0692,
|
| 730 |
+
"step": 10300
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 70.98976109215018,
|
| 734 |
+
"grad_norm": 0.382322758436203,
|
| 735 |
+
"learning_rate": 0.0007757219226004912,
|
| 736 |
+
"loss": 1.0767,
|
| 737 |
+
"step": 10400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 71.67235494880546,
|
| 741 |
+
"grad_norm": 0.3804159462451935,
|
| 742 |
+
"learning_rate": 0.0007713788426715937,
|
| 743 |
+
"loss": 1.0543,
|
| 744 |
+
"step": 10500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 72.35494880546075,
|
| 748 |
+
"grad_norm": 0.3960065245628357,
|
| 749 |
+
"learning_rate": 0.0007670083851991749,
|
| 750 |
+
"loss": 1.0477,
|
| 751 |
+
"step": 10600
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 73.03754266211604,
|
| 755 |
+
"grad_norm": 0.39962372183799744,
|
| 756 |
+
"learning_rate": 0.0007626110906702484,
|
| 757 |
+
"loss": 1.0561,
|
| 758 |
+
"step": 10700
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 73.72013651877133,
|
| 762 |
+
"grad_norm": 0.4019100069999695,
|
| 763 |
+
"learning_rate": 0.0007581875028907215,
|
| 764 |
+
"loss": 1.0329,
|
| 765 |
+
"step": 10800
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 74.40273037542661,
|
| 769 |
+
"grad_norm": 0.42911744117736816,
|
| 770 |
+
"learning_rate": 0.0007537381689181425,
|
| 771 |
+
"loss": 1.0265,
|
| 772 |
+
"step": 10900
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 75.08532423208192,
|
| 776 |
+
"grad_norm": 0.4002211093902588,
|
| 777 |
+
"learning_rate": 0.0007492636389940478,
|
| 778 |
+
"loss": 1.0332,
|
| 779 |
+
"step": 11000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 75.7679180887372,
|
| 783 |
+
"grad_norm": 0.40011659264564514,
|
| 784 |
+
"learning_rate": 0.0007447644664759138,
|
| 785 |
+
"loss": 1.0127,
|
| 786 |
+
"step": 11100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 76.45051194539249,
|
| 790 |
+
"grad_norm": 0.43485236167907715,
|
| 791 |
+
"learning_rate": 0.0007402412077687246,
|
| 792 |
+
"loss": 1.0047,
|
| 793 |
+
"step": 11200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 77.13310580204778,
|
| 797 |
+
"grad_norm": 0.4095554053783417,
|
| 798 |
+
"learning_rate": 0.0007356944222561622,
|
| 799 |
+
"loss": 1.0114,
|
| 800 |
+
"step": 11300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 77.81569965870307,
|
| 804 |
+
"grad_norm": 0.39631539583206177,
|
| 805 |
+
"learning_rate": 0.0007311246722314281,
|
| 806 |
+
"loss": 0.9916,
|
| 807 |
+
"step": 11400
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 78.49829351535836,
|
| 811 |
+
"grad_norm": 0.4598711133003235,
|
| 812 |
+
"learning_rate": 0.0007265325228277065,
|
| 813 |
+
"loss": 0.9822,
|
| 814 |
+
"step": 11500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 79.18088737201366,
|
| 818 |
+
"grad_norm": 0.39505425095558167,
|
| 819 |
+
"learning_rate": 0.0007219185419482739,
|
| 820 |
+
"loss": 0.9883,
|
| 821 |
+
"step": 11600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 79.86348122866895,
|
| 825 |
+
"grad_norm": 0.3844675123691559,
|
| 826 |
+
"learning_rate": 0.0007172833001962683,
|
| 827 |
+
"loss": 0.9701,
|
| 828 |
+
"step": 11700
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 80.54607508532423,
|
| 832 |
+
"grad_norm": 0.48971033096313477,
|
| 833 |
+
"learning_rate": 0.000712627370804124,
|
| 834 |
+
"loss": 0.9597,
|
| 835 |
+
"step": 11800
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 81.22866894197952,
|
| 839 |
+
"grad_norm": 0.40114977955818176,
|
| 840 |
+
"learning_rate": 0.0007079513295626796,
|
| 841 |
+
"loss": 0.9643,
|
| 842 |
+
"step": 11900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 81.91126279863481,
|
| 846 |
+
"grad_norm": 0.42664656043052673,
|
| 847 |
+
"learning_rate": 0.0007032557547499727,
|
| 848 |
+
"loss": 0.9482,
|
| 849 |
+
"step": 12000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 82.5938566552901,
|
| 853 |
+
"grad_norm": 0.5133378505706787,
|
| 854 |
+
"learning_rate": 0.0006985412270597231,
|
| 855 |
+
"loss": 0.9363,
|
| 856 |
+
"step": 12100
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 83.2764505119454,
|
| 860 |
+
"grad_norm": 0.43057283759117126,
|
| 861 |
+
"learning_rate": 0.000693808329529522,
|
| 862 |
+
"loss": 0.9409,
|
| 863 |
+
"step": 12200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 83.95904436860069,
|
| 867 |
+
"grad_norm": 0.4190930426120758,
|
| 868 |
+
"learning_rate": 0.0006890576474687264,
|
| 869 |
+
"loss": 0.9255,
|
| 870 |
+
"step": 12300
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 84.64163822525597,
|
| 874 |
+
"grad_norm": 0.43727907538414,
|
| 875 |
+
"learning_rate": 0.000684289768386076,
|
| 876 |
+
"loss": 0.9119,
|
| 877 |
+
"step": 12400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 85.32423208191126,
|
| 881 |
+
"grad_norm": 0.4994739890098572,
|
| 882 |
+
"learning_rate": 0.000679505281917037,
|
| 883 |
+
"loss": 0.914,
|
| 884 |
+
"step": 12500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 86.00682593856655,
|
| 888 |
+
"grad_norm": 0.45477139949798584,
|
| 889 |
+
"learning_rate": 0.0006747047797508828,
|
| 890 |
+
"loss": 0.9017,
|
| 891 |
+
"step": 12600
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 86.68941979522184,
|
| 895 |
+
"grad_norm": 0.4953708052635193,
|
| 896 |
+
"learning_rate": 0.0006698888555575204,
|
| 897 |
+
"loss": 0.8881,
|
| 898 |
+
"step": 12700
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 87.37201365187714,
|
| 902 |
+
"grad_norm": 0.4449358284473419,
|
| 903 |
+
"learning_rate": 0.000665058104914073,
|
| 904 |
+
"loss": 0.8875,
|
| 905 |
+
"step": 12800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 88.05460750853243,
|
| 909 |
+
"grad_norm": 0.4819263815879822,
|
| 910 |
+
"learning_rate": 0.0006602131252312249,
|
| 911 |
+
"loss": 0.8748,
|
| 912 |
+
"step": 12900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 88.73720136518772,
|
| 916 |
+
"grad_norm": 0.5412826538085938,
|
| 917 |
+
"learning_rate": 0.0006553545156793422,
|
| 918 |
+
"loss": 0.8655,
|
| 919 |
+
"step": 13000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 89.419795221843,
|
| 923 |
+
"grad_norm": 0.4753580391407013,
|
| 924 |
+
"learning_rate": 0.0006504828771143733,
|
| 925 |
+
"loss": 0.8617,
|
| 926 |
+
"step": 13100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 90.10238907849829,
|
| 930 |
+
"grad_norm": 0.545566737651825,
|
| 931 |
+
"learning_rate": 0.0006455988120035418,
|
| 932 |
+
"loss": 0.85,
|
| 933 |
+
"step": 13200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 90.78498293515358,
|
| 937 |
+
"grad_norm": 0.5255779027938843,
|
| 938 |
+
"learning_rate": 0.0006407029243508423,
|
| 939 |
+
"loss": 0.8449,
|
| 940 |
+
"step": 13300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 91.46757679180887,
|
| 944 |
+
"grad_norm": 0.43862029910087585,
|
| 945 |
+
"learning_rate": 0.0006357958196223419,
|
| 946 |
+
"loss": 0.8371,
|
| 947 |
+
"step": 13400
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 92.15017064846417,
|
| 951 |
+
"grad_norm": 0.5360426902770996,
|
| 952 |
+
"learning_rate": 0.0006308781046713046,
|
| 953 |
+
"loss": 0.8257,
|
| 954 |
+
"step": 13500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 92.83276450511946,
|
| 958 |
+
"grad_norm": 0.5451890230178833,
|
| 959 |
+
"learning_rate": 0.0006259503876631428,
|
| 960 |
+
"loss": 0.8247,
|
| 961 |
+
"step": 13600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 93.51535836177474,
|
| 965 |
+
"grad_norm": 0.4524667263031006,
|
| 966 |
+
"learning_rate": 0.0006210132780002057,
|
| 967 |
+
"loss": 0.8147,
|
| 968 |
+
"step": 13700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 94.19795221843003,
|
| 972 |
+
"grad_norm": 0.6203665733337402,
|
| 973 |
+
"learning_rate": 0.0006160673862464167,
|
| 974 |
+
"loss": 0.8022,
|
| 975 |
+
"step": 13800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 94.88054607508532,
|
| 979 |
+
"grad_norm": 0.5576283931732178,
|
| 980 |
+
"learning_rate": 0.0006111133240517648,
|
| 981 |
+
"loss": 0.8057,
|
| 982 |
+
"step": 13900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 95.56313993174061,
|
| 986 |
+
"grad_norm": 0.48380613327026367,
|
| 987 |
+
"learning_rate": 0.000606151704076664,
|
| 988 |
+
"loss": 0.7906,
|
| 989 |
+
"step": 14000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 96.24573378839591,
|
| 993 |
+
"grad_norm": 0.5328605771064758,
|
| 994 |
+
"learning_rate": 0.0006011831399161856,
|
| 995 |
+
"loss": 0.7823,
|
| 996 |
+
"step": 14100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 96.9283276450512,
|
| 1000 |
+
"grad_norm": 0.5395786762237549,
|
| 1001 |
+
"learning_rate": 0.0005962082460241767,
|
| 1002 |
+
"loss": 0.7883,
|
| 1003 |
+
"step": 14200
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 97.61092150170649,
|
| 1007 |
+
"grad_norm": 0.48329558968544006,
|
| 1008 |
+
"learning_rate": 0.0005912276376372715,
|
| 1009 |
+
"loss": 0.7707,
|
| 1010 |
+
"step": 14300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 98.29351535836177,
|
| 1014 |
+
"grad_norm": 0.5496215224266052,
|
| 1015 |
+
"learning_rate": 0.0005862419306988061,
|
| 1016 |
+
"loss": 0.7636,
|
| 1017 |
+
"step": 14400
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 98.97610921501706,
|
| 1021 |
+
"grad_norm": 0.6007915139198303,
|
| 1022 |
+
"learning_rate": 0.0005812517417826456,
|
| 1023 |
+
"loss": 0.7712,
|
| 1024 |
+
"step": 14500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 99.65870307167235,
|
| 1028 |
+
"grad_norm": 0.4981062710285187,
|
| 1029 |
+
"learning_rate": 0.000576257688016934,
|
| 1030 |
+
"loss": 0.7529,
|
| 1031 |
+
"step": 14600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 100.34129692832765,
|
| 1035 |
+
"grad_norm": 0.5475151538848877,
|
| 1036 |
+
"learning_rate": 0.0005712603870077736,
|
| 1037 |
+
"loss": 0.7468,
|
| 1038 |
+
"step": 14700
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 101.02389078498294,
|
| 1042 |
+
"grad_norm": 0.4879900813102722,
|
| 1043 |
+
"learning_rate": 0.0005662604567628494,
|
| 1044 |
+
"loss": 0.7568,
|
| 1045 |
+
"step": 14800
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 101.70648464163823,
|
| 1049 |
+
"grad_norm": 0.4963570535182953,
|
| 1050 |
+
"learning_rate": 0.0005612585156149982,
|
| 1051 |
+
"loss": 0.7386,
|
| 1052 |
+
"step": 14900
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 102.38907849829351,
|
| 1056 |
+
"grad_norm": 0.5414558053016663,
|
| 1057 |
+
"learning_rate": 0.0005562551821457418,
|
| 1058 |
+
"loss": 0.7326,
|
| 1059 |
+
"step": 15000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 103.0716723549488,
|
| 1063 |
+
"grad_norm": 0.47427278757095337,
|
| 1064 |
+
"learning_rate": 0.0005512510751087889,
|
| 1065 |
+
"loss": 0.7415,
|
| 1066 |
+
"step": 15100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 103.75426621160409,
|
| 1070 |
+
"grad_norm": 0.48589903116226196,
|
| 1071 |
+
"learning_rate": 0.0005462468133535131,
|
| 1072 |
+
"loss": 0.7267,
|
| 1073 |
+
"step": 15200
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 104.43686006825939,
|
| 1077 |
+
"grad_norm": 0.5541400909423828,
|
| 1078 |
+
"learning_rate": 0.0005412430157484224,
|
| 1079 |
+
"loss": 0.7206,
|
| 1080 |
+
"step": 15300
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 105.11945392491468,
|
| 1084 |
+
"grad_norm": 0.4941166341304779,
|
| 1085 |
+
"learning_rate": 0.0005362403011046234,
|
| 1086 |
+
"loss": 0.7289,
|
| 1087 |
+
"step": 15400
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 105.80204778156997,
|
| 1091 |
+
"grad_norm": 0.4855073094367981,
|
| 1092 |
+
"learning_rate": 0.0005312392880992954,
|
| 1093 |
+
"loss": 0.716,
|
| 1094 |
+
"step": 15500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 106.48464163822526,
|
| 1098 |
+
"grad_norm": 0.5357630252838135,
|
| 1099 |
+
"learning_rate": 0.0005262405951991788,
|
| 1100 |
+
"loss": 0.7111,
|
| 1101 |
+
"step": 15600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 107.16723549488054,
|
| 1105 |
+
"grad_norm": 0.5147479772567749,
|
| 1106 |
+
"learning_rate": 0.0005212448405840905,
|
| 1107 |
+
"loss": 0.7181,
|
| 1108 |
+
"step": 15700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 107.84982935153583,
|
| 1112 |
+
"grad_norm": 0.47526028752326965,
|
| 1113 |
+
"learning_rate": 0.0005162526420704744,
|
| 1114 |
+
"loss": 0.7072,
|
| 1115 |
+
"step": 15800
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 108.53242320819113,
|
| 1119 |
+
"grad_norm": 0.528582751750946,
|
| 1120 |
+
"learning_rate": 0.0005112646170349981,
|
| 1121 |
+
"loss": 0.7021,
|
| 1122 |
+
"step": 15900
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 109.21501706484642,
|
| 1126 |
+
"grad_norm": 0.4804699420928955,
|
| 1127 |
+
"learning_rate": 0.0005062813823382014,
|
| 1128 |
+
"loss": 0.7093,
|
| 1129 |
+
"step": 16000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 109.89761092150171,
|
| 1133 |
+
"grad_norm": 0.5126569271087646,
|
| 1134 |
+
"learning_rate": 0.000501303554248212,
|
| 1135 |
+
"loss": 0.7008,
|
| 1136 |
+
"step": 16100
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 110.580204778157,
|
| 1140 |
+
"grad_norm": 0.5352160930633545,
|
| 1141 |
+
"learning_rate": 0.0004963317483645311,
|
| 1142 |
+
"loss": 0.6946,
|
| 1143 |
+
"step": 16200
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 111.26279863481228,
|
| 1147 |
+
"grad_norm": 0.5465342402458191,
|
| 1148 |
+
"learning_rate": 0.0004913665795419041,
|
| 1149 |
+
"loss": 0.7008,
|
| 1150 |
+
"step": 16300
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 111.94539249146757,
|
| 1154 |
+
"grad_norm": 0.45367056131362915,
|
| 1155 |
+
"learning_rate": 0.00048640866181428265,
|
| 1156 |
+
"loss": 0.6939,
|
| 1157 |
+
"step": 16400
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 112.62798634812286,
|
| 1161 |
+
"grad_norm": 0.6308451294898987,
|
| 1162 |
+
"learning_rate": 0.00048145860831888776,
|
| 1163 |
+
"loss": 0.6885,
|
| 1164 |
+
"step": 16500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 113.31058020477816,
|
| 1168 |
+
"grad_norm": 0.4779800772666931,
|
| 1169 |
+
"learning_rate": 0.00047651703122038395,
|
| 1170 |
+
"loss": 0.6939,
|
| 1171 |
+
"step": 16600
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 113.99317406143345,
|
| 1175 |
+
"grad_norm": 0.4980040192604065,
|
| 1176 |
+
"learning_rate": 0.0004715845416351744,
|
| 1177 |
+
"loss": 0.6879,
|
| 1178 |
+
"step": 16700
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 114.67576791808874,
|
| 1182 |
+
"grad_norm": 0.5260293483734131,
|
| 1183 |
+
"learning_rate": 0.00046666174955582507,
|
| 1184 |
+
"loss": 0.6835,
|
| 1185 |
+
"step": 16800
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 115.35836177474403,
|
| 1189 |
+
"grad_norm": 0.47831013798713684,
|
| 1190 |
+
"learning_rate": 0.0004617492637756282,
|
| 1191 |
+
"loss": 0.6877,
|
| 1192 |
+
"step": 16900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 116.04095563139931,
|
| 1196 |
+
"grad_norm": 0.5765718221664429,
|
| 1197 |
+
"learning_rate": 0.0004568476918133129,
|
| 1198 |
+
"loss": 0.6813,
|
| 1199 |
+
"step": 17000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 116.7235494880546,
|
| 1203 |
+
"grad_norm": 0.5322794318199158,
|
| 1204 |
+
"learning_rate": 0.00045195763983791573,
|
| 1205 |
+
"loss": 0.679,
|
| 1206 |
+
"step": 17100
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 117.4061433447099,
|
| 1210 |
+
"grad_norm": 0.514970064163208,
|
| 1211 |
+
"learning_rate": 0.00044707971259381584,
|
| 1212 |
+
"loss": 0.68,
|
| 1213 |
+
"step": 17200
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 118.08873720136519,
|
| 1217 |
+
"grad_norm": 0.539171576499939,
|
| 1218 |
+
"learning_rate": 0.00044221451332594785,
|
| 1219 |
+
"loss": 0.6769,
|
| 1220 |
+
"step": 17300
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 118.77133105802048,
|
| 1224 |
+
"grad_norm": 0.5568128824234009,
|
| 1225 |
+
"learning_rate": 0.00043736264370519863,
|
| 1226 |
+
"loss": 0.6757,
|
| 1227 |
+
"step": 17400
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 119.45392491467577,
|
| 1231 |
+
"grad_norm": 0.4861965477466583,
|
| 1232 |
+
"learning_rate": 0.0004325247037540011,
|
| 1233 |
+
"loss": 0.6749,
|
| 1234 |
+
"step": 17500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 120.13651877133105,
|
| 1238 |
+
"grad_norm": 0.49631592631340027,
|
| 1239 |
+
"learning_rate": 0.00042770129177212914,
|
| 1240 |
+
"loss": 0.671,
|
| 1241 |
+
"step": 17600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 120.81911262798634,
|
| 1245 |
+
"grad_norm": 0.5209415555000305,
|
| 1246 |
+
"learning_rate": 0.00042289300426270806,
|
| 1247 |
+
"loss": 0.6714,
|
| 1248 |
+
"step": 17700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 121.50170648464164,
|
| 1252 |
+
"grad_norm": 0.4455106556415558,
|
| 1253 |
+
"learning_rate": 0.0004181004358584447,
|
| 1254 |
+
"loss": 0.6699,
|
| 1255 |
+
"step": 17800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 122.18430034129693,
|
| 1259 |
+
"grad_norm": 0.46950340270996094,
|
| 1260 |
+
"learning_rate": 0.00041332417924809175,
|
| 1261 |
+
"loss": 0.6654,
|
| 1262 |
+
"step": 17900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 122.86689419795222,
|
| 1266 |
+
"grad_norm": 0.5346481800079346,
|
| 1267 |
+
"learning_rate": 0.00040856482510314985,
|
| 1268 |
+
"loss": 0.6681,
|
| 1269 |
+
"step": 18000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 123.54948805460751,
|
| 1273 |
+
"grad_norm": 0.47252893447875977,
|
| 1274 |
+
"learning_rate": 0.0004038229620048207,
|
| 1275 |
+
"loss": 0.6648,
|
| 1276 |
+
"step": 18100
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 124.2320819112628,
|
| 1280 |
+
"grad_norm": 0.5372588634490967,
|
| 1281 |
+
"learning_rate": 0.00039909917637121796,
|
| 1282 |
+
"loss": 0.6611,
|
| 1283 |
+
"step": 18200
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 124.91467576791808,
|
| 1287 |
+
"grad_norm": 0.5461378693580627,
|
| 1288 |
+
"learning_rate": 0.00039439405238484614,
|
| 1289 |
+
"loss": 0.6637,
|
| 1290 |
+
"step": 18300
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 125.59726962457339,
|
| 1294 |
+
"grad_norm": 0.4744736850261688,
|
| 1295 |
+
"learning_rate": 0.0003897081719203558,
|
| 1296 |
+
"loss": 0.659,
|
| 1297 |
+
"step": 18400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 126.27986348122867,
|
| 1301 |
+
"grad_norm": 0.5168038606643677,
|
| 1302 |
+
"learning_rate": 0.0003850421144725844,
|
| 1303 |
+
"loss": 0.6559,
|
| 1304 |
+
"step": 18500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 126.96245733788396,
|
| 1308 |
+
"grad_norm": 0.5486589670181274,
|
| 1309 |
+
"learning_rate": 0.00038039645708488955,
|
| 1310 |
+
"loss": 0.6606,
|
| 1311 |
+
"step": 18600
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 127.64505119453925,
|
| 1315 |
+
"grad_norm": 0.44690069556236267,
|
| 1316 |
+
"learning_rate": 0.0003757717742777895,
|
| 1317 |
+
"loss": 0.6538,
|
| 1318 |
+
"step": 18700
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 128.32764505119454,
|
| 1322 |
+
"grad_norm": 0.5112203359603882,
|
| 1323 |
+
"learning_rate": 0.00037116863797791145,
|
| 1324 |
+
"loss": 0.6523,
|
| 1325 |
+
"step": 18800
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 129.01023890784984,
|
| 1329 |
+
"grad_norm": 0.479837030172348,
|
| 1330 |
+
"learning_rate": 0.0003665876174472628,
|
| 1331 |
+
"loss": 0.6563,
|
| 1332 |
+
"step": 18900
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 129.6928327645051,
|
| 1336 |
+
"grad_norm": 0.4698416292667389,
|
| 1337 |
+
"learning_rate": 0.00036202927921283144,
|
| 1338 |
+
"loss": 0.6495,
|
| 1339 |
+
"step": 19000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 130.37542662116041,
|
| 1343 |
+
"grad_norm": 0.5283064246177673,
|
| 1344 |
+
"learning_rate": 0.0003574941869965248,
|
| 1345 |
+
"loss": 0.6474,
|
| 1346 |
+
"step": 19100
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 131.0580204778157,
|
| 1350 |
+
"grad_norm": 0.4855287969112396,
|
| 1351 |
+
"learning_rate": 0.00035298290164545443,
|
| 1352 |
+
"loss": 0.652,
|
| 1353 |
+
"step": 19200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 131.740614334471,
|
| 1357 |
+
"grad_norm": 0.4482058882713318,
|
| 1358 |
+
"learning_rate": 0.00034849598106257795,
|
| 1359 |
+
"loss": 0.6454,
|
| 1360 |
+
"step": 19300
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 132.4232081911263,
|
| 1364 |
+
"grad_norm": 0.487212210893631,
|
| 1365 |
+
"learning_rate": 0.00034403398013770347,
|
| 1366 |
+
"loss": 0.643,
|
| 1367 |
+
"step": 19400
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 133.10580204778157,
|
| 1371 |
+
"grad_norm": 0.4655930697917938,
|
| 1372 |
+
"learning_rate": 0.0003395974506788674,
|
| 1373 |
+
"loss": 0.6477,
|
| 1374 |
+
"step": 19500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 133.78839590443687,
|
| 1378 |
+
"grad_norm": 0.4315454661846161,
|
| 1379 |
+
"learning_rate": 0.00033518694134409367,
|
| 1380 |
+
"loss": 0.6411,
|
| 1381 |
+
"step": 19600
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 134.47098976109214,
|
| 1385 |
+
"grad_norm": 0.505107581615448,
|
| 1386 |
+
"learning_rate": 0.00033080299757354216,
|
| 1387 |
+
"loss": 0.6396,
|
| 1388 |
+
"step": 19700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 135.15358361774744,
|
| 1392 |
+
"grad_norm": 0.47746554017066956,
|
| 1393 |
+
"learning_rate": 0.00032644616152205385,
|
| 1394 |
+
"loss": 0.6436,
|
| 1395 |
+
"step": 19800
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 135.83617747440272,
|
| 1399 |
+
"grad_norm": 0.44861286878585815,
|
| 1400 |
+
"learning_rate": 0.0003221169719921052,
|
| 1401 |
+
"loss": 0.6382,
|
| 1402 |
+
"step": 19900
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 136.51877133105802,
|
| 1406 |
+
"grad_norm": 0.4924197196960449,
|
| 1407 |
+
"learning_rate": 0.00031781596436717465,
|
| 1408 |
+
"loss": 0.636,
|
| 1409 |
+
"step": 20000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 137.20136518771332,
|
| 1413 |
+
"grad_norm": 0.4533005952835083,
|
| 1414 |
+
"learning_rate": 0.0003135436705455319,
|
| 1415 |
+
"loss": 0.6388,
|
| 1416 |
+
"step": 20100
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 137.8839590443686,
|
| 1420 |
+
"grad_norm": 0.4372631311416626,
|
| 1421 |
+
"learning_rate": 0.0003093006188744598,
|
| 1422 |
+
"loss": 0.6349,
|
| 1423 |
+
"step": 20200
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 138.5665529010239,
|
| 1427 |
+
"grad_norm": 0.5831769108772278,
|
| 1428 |
+
"learning_rate": 0.0003050873340849148,
|
| 1429 |
+
"loss": 0.6323,
|
| 1430 |
+
"step": 20300
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 139.24914675767917,
|
| 1434 |
+
"grad_norm": 0.41959720849990845,
|
| 1435 |
+
"learning_rate": 0.00030090433722663367,
|
| 1436 |
+
"loss": 0.6357,
|
| 1437 |
+
"step": 20400
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 139.93174061433447,
|
| 1441 |
+
"grad_norm": 0.47732993960380554,
|
| 1442 |
+
"learning_rate": 0.00029675214560369637,
|
| 1443 |
+
"loss": 0.6323,
|
| 1444 |
+
"step": 20500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 140.61433447098977,
|
| 1448 |
+
"grad_norm": 0.4794413447380066,
|
| 1449 |
+
"learning_rate": 0.0002926312727105518,
|
| 1450 |
+
"loss": 0.6303,
|
| 1451 |
+
"step": 20600
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 141.29692832764505,
|
| 1455 |
+
"grad_norm": 0.42916139960289,
|
| 1456 |
+
"learning_rate": 0.00028854222816851547,
|
| 1457 |
+
"loss": 0.6327,
|
| 1458 |
+
"step": 20700
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 141.97952218430035,
|
| 1462 |
+
"grad_norm": 0.42981767654418945,
|
| 1463 |
+
"learning_rate": 0.00028448551766274466,
|
| 1464 |
+
"loss": 0.6285,
|
| 1465 |
+
"step": 20800
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 142.66211604095562,
|
| 1469 |
+
"grad_norm": 0.48353445529937744,
|
| 1470 |
+
"learning_rate": 0.0002804616428797021,
|
| 1471 |
+
"loss": 0.6272,
|
| 1472 |
+
"step": 20900
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 143.34470989761093,
|
| 1476 |
+
"grad_norm": 0.4736007750034332,
|
| 1477 |
+
"learning_rate": 0.0002764711014451121,
|
| 1478 |
+
"loss": 0.6288,
|
| 1479 |
+
"step": 21000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 144.0273037542662,
|
| 1483 |
+
"grad_norm": 0.47045692801475525,
|
| 1484 |
+
"learning_rate": 0.0002725143868624209,
|
| 1485 |
+
"loss": 0.6257,
|
| 1486 |
+
"step": 21100
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 144.7098976109215,
|
| 1490 |
+
"grad_norm": 0.4963953495025635,
|
| 1491 |
+
"learning_rate": 0.0002685919884517659,
|
| 1492 |
+
"loss": 0.6251,
|
| 1493 |
+
"step": 21200
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 145.3924914675768,
|
| 1497 |
+
"grad_norm": 0.4388594925403595,
|
| 1498 |
+
"learning_rate": 0.00026470439128946214,
|
| 1499 |
+
"loss": 0.6249,
|
| 1500 |
+
"step": 21300
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 146.07508532423208,
|
| 1504 |
+
"grad_norm": 0.4757547378540039,
|
| 1505 |
+
"learning_rate": 0.0002608520761480135,
|
| 1506 |
+
"loss": 0.6241,
|
| 1507 |
+
"step": 21400
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 146.75767918088738,
|
| 1511 |
+
"grad_norm": 0.46713003516197205,
|
| 1512 |
+
"learning_rate": 0.00025703551943665636,
|
| 1513 |
+
"loss": 0.6229,
|
| 1514 |
+
"step": 21500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 147.44027303754265,
|
| 1518 |
+
"grad_norm": 0.43607574701309204,
|
| 1519 |
+
"learning_rate": 0.00025325519314244374,
|
| 1520 |
+
"loss": 0.623,
|
| 1521 |
+
"step": 21600
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 148.12286689419795,
|
| 1525 |
+
"grad_norm": 0.4766788184642792,
|
| 1526 |
+
"learning_rate": 0.00024951156477187413,
|
| 1527 |
+
"loss": 0.6212,
|
| 1528 |
+
"step": 21700
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 148.80546075085323,
|
| 1532 |
+
"grad_norm": 0.49822133779525757,
|
| 1533 |
+
"learning_rate": 0.00024580509729307673,
|
| 1534 |
+
"loss": 0.6215,
|
| 1535 |
+
"step": 21800
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 149.48805460750853,
|
| 1539 |
+
"grad_norm": 0.4069366455078125,
|
| 1540 |
+
"learning_rate": 0.0002421362490785569,
|
| 1541 |
+
"loss": 0.6199,
|
| 1542 |
+
"step": 21900
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 150.17064846416383,
|
| 1546 |
+
"grad_norm": 0.46552979946136475,
|
| 1547 |
+
"learning_rate": 0.00023850547384850974,
|
| 1548 |
+
"loss": 0.6194,
|
| 1549 |
+
"step": 22000
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 150.8532423208191,
|
| 1553 |
+
"grad_norm": 0.45079755783081055,
|
| 1554 |
+
"learning_rate": 0.00023491322061470915,
|
| 1555 |
+
"loss": 0.6202,
|
| 1556 |
+
"step": 22100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 151.5358361774744,
|
| 1560 |
+
"grad_norm": 0.46759602427482605,
|
| 1561 |
+
"learning_rate": 0.00023135993362497948,
|
| 1562 |
+
"loss": 0.6181,
|
| 1563 |
+
"step": 22200
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 152.21843003412968,
|
| 1567 |
+
"grad_norm": 0.45771974325180054,
|
| 1568 |
+
"learning_rate": 0.00022784605230825666,
|
| 1569 |
+
"loss": 0.6164,
|
| 1570 |
+
"step": 22300
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 152.90102389078498,
|
| 1574 |
+
"grad_norm": 0.5069631338119507,
|
| 1575 |
+
"learning_rate": 0.00022437201122024375,
|
| 1576 |
+
"loss": 0.6187,
|
| 1577 |
+
"step": 22400
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 153.5836177474403,
|
| 1581 |
+
"grad_norm": 0.43779122829437256,
|
| 1582 |
+
"learning_rate": 0.00022093823998967115,
|
| 1583 |
+
"loss": 0.6164,
|
| 1584 |
+
"step": 22500
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 154.26621160409556,
|
| 1588 |
+
"grad_norm": 0.477224200963974,
|
| 1589 |
+
"learning_rate": 0.00021754516326516458,
|
| 1590 |
+
"loss": 0.6151,
|
| 1591 |
+
"step": 22600
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 154.94880546075086,
|
| 1595 |
+
"grad_norm": 0.47925087809562683,
|
| 1596 |
+
"learning_rate": 0.00021419320066272973,
|
| 1597 |
+
"loss": 0.6166,
|
| 1598 |
+
"step": 22700
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 155.63139931740614,
|
| 1602 |
+
"grad_norm": 0.4455689787864685,
|
| 1603 |
+
"learning_rate": 0.00021088276671385882,
|
| 1604 |
+
"loss": 0.6139,
|
| 1605 |
+
"step": 22800
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 156.31399317406144,
|
| 1609 |
+
"grad_norm": 0.4499022364616394,
|
| 1610 |
+
"learning_rate": 0.00020761427081426677,
|
| 1611 |
+
"loss": 0.6132,
|
| 1612 |
+
"step": 22900
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 156.9965870307167,
|
| 1616 |
+
"grad_norm": 0.4519612193107605,
|
| 1617 |
+
"learning_rate": 0.00020438811717326112,
|
| 1618 |
+
"loss": 0.6154,
|
| 1619 |
+
"step": 23000
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 157.679180887372,
|
| 1623 |
+
"grad_norm": 0.4130764603614807,
|
| 1624 |
+
"learning_rate": 0.0002012047047637547,
|
| 1625 |
+
"loss": 0.6122,
|
| 1626 |
+
"step": 23100
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 158.36177474402731,
|
| 1630 |
+
"grad_norm": 0.44488415122032166,
|
| 1631 |
+
"learning_rate": 0.0001980644272729254,
|
| 1632 |
+
"loss": 0.6132,
|
| 1633 |
+
"step": 23200
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 159.0443686006826,
|
| 1637 |
+
"grad_norm": 0.47042757272720337,
|
| 1638 |
+
"learning_rate": 0.00019496767305352946,
|
| 1639 |
+
"loss": 0.6135,
|
| 1640 |
+
"step": 23300
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 159.7269624573379,
|
| 1644 |
+
"grad_norm": 0.3974725604057312,
|
| 1645 |
+
"learning_rate": 0.00019191482507587376,
|
| 1646 |
+
"loss": 0.6107,
|
| 1647 |
+
"step": 23400
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 160.40955631399316,
|
| 1651 |
+
"grad_norm": 0.46765318512916565,
|
| 1652 |
+
"learning_rate": 0.00018890626088045593,
|
| 1653 |
+
"loss": 0.6109,
|
| 1654 |
+
"step": 23500
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 161.09215017064847,
|
| 1658 |
+
"grad_norm": 0.42309921979904175,
|
| 1659 |
+
"learning_rate": 0.00018594235253127373,
|
| 1660 |
+
"loss": 0.6122,
|
| 1661 |
+
"step": 23600
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 161.77474402730377,
|
| 1665 |
+
"grad_norm": 0.3989265561103821,
|
| 1666 |
+
"learning_rate": 0.00018302346656981234,
|
| 1667 |
+
"loss": 0.6099,
|
| 1668 |
+
"step": 23700
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 162.45733788395904,
|
| 1672 |
+
"grad_norm": 0.4663792550563812,
|
| 1673 |
+
"learning_rate": 0.0001801499639697153,
|
| 1674 |
+
"loss": 0.6103,
|
| 1675 |
+
"step": 23800
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 163.13993174061434,
|
| 1679 |
+
"grad_norm": 0.3913436233997345,
|
| 1680 |
+
"learning_rate": 0.0001773222000921435,
|
| 1681 |
+
"loss": 0.6106,
|
| 1682 |
+
"step": 23900
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 163.82252559726962,
|
| 1686 |
+
"grad_norm": 0.41263365745544434,
|
| 1687 |
+
"learning_rate": 0.00017454052464182778,
|
| 1688 |
+
"loss": 0.6092,
|
| 1689 |
+
"step": 24000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 164.50511945392492,
|
| 1693 |
+
"grad_norm": 0.4633448123931885,
|
| 1694 |
+
"learning_rate": 0.0001718052816238222,
|
| 1695 |
+
"loss": 0.608,
|
| 1696 |
+
"step": 24100
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 165.1877133105802,
|
| 1700 |
+
"grad_norm": 0.3786383271217346,
|
| 1701 |
+
"learning_rate": 0.00016911680930096085,
|
| 1702 |
+
"loss": 0.6094,
|
| 1703 |
+
"step": 24200
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 165.8703071672355,
|
| 1707 |
+
"grad_norm": 0.4065646529197693,
|
| 1708 |
+
"learning_rate": 0.00016647544015202623,
|
| 1709 |
+
"loss": 0.608,
|
| 1710 |
+
"step": 24300
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 166.5529010238908,
|
| 1714 |
+
"grad_norm": 0.43729546666145325,
|
| 1715 |
+
"learning_rate": 0.00016388150083063117,
|
| 1716 |
+
"loss": 0.6072,
|
| 1717 |
+
"step": 24400
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 167.23549488054607,
|
| 1721 |
+
"grad_norm": 0.4111308157444,
|
| 1722 |
+
"learning_rate": 0.00016133531212482325,
|
| 1723 |
+
"loss": 0.6078,
|
| 1724 |
+
"step": 24500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 167.91808873720137,
|
| 1728 |
+
"grad_norm": 0.39553767442703247,
|
| 1729 |
+
"learning_rate": 0.00015883718891741258,
|
| 1730 |
+
"loss": 0.6066,
|
| 1731 |
+
"step": 24600
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 168.60068259385665,
|
| 1735 |
+
"grad_norm": 0.4074500501155853,
|
| 1736 |
+
"learning_rate": 0.00015638744014703128,
|
| 1737 |
+
"loss": 0.6068,
|
| 1738 |
+
"step": 24700
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 169.28327645051195,
|
| 1742 |
+
"grad_norm": 0.42327943444252014,
|
| 1743 |
+
"learning_rate": 0.0001539863687699274,
|
| 1744 |
+
"loss": 0.607,
|
| 1745 |
+
"step": 24800
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 169.96587030716722,
|
| 1749 |
+
"grad_norm": 0.40863099694252014,
|
| 1750 |
+
"learning_rate": 0.0001516342717224992,
|
| 1751 |
+
"loss": 0.6067,
|
| 1752 |
+
"step": 24900
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 170.64846416382252,
|
| 1756 |
+
"grad_norm": 0.4853709638118744,
|
| 1757 |
+
"learning_rate": 0.00014933143988457237,
|
| 1758 |
+
"loss": 0.6056,
|
| 1759 |
+
"step": 25000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 171.33105802047783,
|
| 1763 |
+
"grad_norm": 0.383801132440567,
|
| 1764 |
+
"learning_rate": 0.00014707815804342907,
|
| 1765 |
+
"loss": 0.6063,
|
| 1766 |
+
"step": 25100
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 172.0136518771331,
|
| 1770 |
+
"grad_norm": 0.4241873621940613,
|
| 1771 |
+
"learning_rate": 0.00014487470485858767,
|
| 1772 |
+
"loss": 0.6058,
|
| 1773 |
+
"step": 25200
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 172.6962457337884,
|
| 1777 |
+
"grad_norm": 0.4262351989746094,
|
| 1778 |
+
"learning_rate": 0.00014272135282734167,
|
| 1779 |
+
"loss": 0.6043,
|
| 1780 |
+
"step": 25300
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 173.37883959044368,
|
| 1784 |
+
"grad_norm": 0.38850197196006775,
|
| 1785 |
+
"learning_rate": 0.00014061836825106022,
|
| 1786 |
+
"loss": 0.606,
|
| 1787 |
+
"step": 25400
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 174.06143344709898,
|
| 1791 |
+
"grad_norm": 0.4169691801071167,
|
| 1792 |
+
"learning_rate": 0.00013856601120225574,
|
| 1793 |
+
"loss": 0.6038,
|
| 1794 |
+
"step": 25500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 174.74402730375428,
|
| 1798 |
+
"grad_norm": 0.43140584230422974,
|
| 1799 |
+
"learning_rate": 0.00013656453549242055,
|
| 1800 |
+
"loss": 0.6043,
|
| 1801 |
+
"step": 25600
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 175.42662116040955,
|
| 1805 |
+
"grad_norm": 0.3729684054851532,
|
| 1806 |
+
"learning_rate": 0.00013461418864063861,
|
| 1807 |
+
"loss": 0.6047,
|
| 1808 |
+
"step": 25700
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 176.10921501706486,
|
| 1812 |
+
"grad_norm": 0.435611367225647,
|
| 1813 |
+
"learning_rate": 0.00013271521184297523,
|
| 1814 |
+
"loss": 0.6035,
|
| 1815 |
+
"step": 25800
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 176.79180887372013,
|
| 1819 |
+
"grad_norm": 0.4359755218029022,
|
| 1820 |
+
"learning_rate": 0.00013086783994264905,
|
| 1821 |
+
"loss": 0.6044,
|
| 1822 |
+
"step": 25900
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 177.47440273037543,
|
| 1826 |
+
"grad_norm": 0.3867483139038086,
|
| 1827 |
+
"learning_rate": 0.00012907230140098895,
|
| 1828 |
+
"loss": 0.6031,
|
| 1829 |
+
"step": 26000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 178.1569965870307,
|
| 1833 |
+
"grad_norm": 0.41799867153167725,
|
| 1834 |
+
"learning_rate": 0.00012732881826918094,
|
| 1835 |
+
"loss": 0.6031,
|
| 1836 |
+
"step": 26100
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 178.839590443686,
|
| 1840 |
+
"grad_norm": 0.43031203746795654,
|
| 1841 |
+
"learning_rate": 0.00012563760616080745,
|
| 1842 |
+
"loss": 0.6028,
|
| 1843 |
+
"step": 26200
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 179.5221843003413,
|
| 1847 |
+
"grad_norm": 0.3603518605232239,
|
| 1848 |
+
"learning_rate": 0.00012399887422518243,
|
| 1849 |
+
"loss": 0.6026,
|
| 1850 |
+
"step": 26300
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 180.20477815699658,
|
| 1854 |
+
"grad_norm": 0.42124807834625244,
|
| 1855 |
+
"learning_rate": 0.0001224128251214867,
|
| 1856 |
+
"loss": 0.6021,
|
| 1857 |
+
"step": 26400
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 180.88737201365188,
|
| 1861 |
+
"grad_norm": 0.40946000814437866,
|
| 1862 |
+
"learning_rate": 0.00012087965499370513,
|
| 1863 |
+
"loss": 0.6036,
|
| 1864 |
+
"step": 26500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 181.56996587030716,
|
| 1868 |
+
"grad_norm": 0.3763185143470764,
|
| 1869 |
+
"learning_rate": 0.00011939955344636983,
|
| 1870 |
+
"loss": 0.6022,
|
| 1871 |
+
"step": 26600
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 182.25255972696246,
|
| 1875 |
+
"grad_norm": 0.4167167544364929,
|
| 1876 |
+
"learning_rate": 0.00011797270352111224,
|
| 1877 |
+
"loss": 0.6016,
|
| 1878 |
+
"step": 26700
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 182.93515358361773,
|
| 1882 |
+
"grad_norm": 0.42429491877555847,
|
| 1883 |
+
"learning_rate": 0.0001165992816740268,
|
| 1884 |
+
"loss": 0.6021,
|
| 1885 |
+
"step": 26800
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 183.61774744027304,
|
| 1889 |
+
"grad_norm": 0.38188785314559937,
|
| 1890 |
+
"learning_rate": 0.00011527945775384831,
|
| 1891 |
+
"loss": 0.6022,
|
| 1892 |
+
"step": 26900
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 184.30034129692834,
|
| 1896 |
+
"grad_norm": 0.47042107582092285,
|
| 1897 |
+
"learning_rate": 0.00011401339498094774,
|
| 1898 |
+
"loss": 0.6016,
|
| 1899 |
+
"step": 27000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 184.9829351535836,
|
| 1903 |
+
"grad_norm": 0.422690212726593,
|
| 1904 |
+
"learning_rate": 0.00011280124992714677,
|
| 1905 |
+
"loss": 0.602,
|
| 1906 |
+
"step": 27100
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 185.6655290102389,
|
| 1910 |
+
"grad_norm": 0.3714023232460022,
|
| 1911 |
+
"learning_rate": 0.00011164317249635475,
|
| 1912 |
+
"loss": 0.6014,
|
| 1913 |
+
"step": 27200
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 186.3481228668942,
|
| 1917 |
+
"grad_norm": 0.43933653831481934,
|
| 1918 |
+
"learning_rate": 0.00011053930590603041,
|
| 1919 |
+
"loss": 0.6003,
|
| 1920 |
+
"step": 27300
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 187.0307167235495,
|
| 1924 |
+
"grad_norm": 0.3583015501499176,
|
| 1925 |
+
"learning_rate": 0.00010948978666947039,
|
| 1926 |
+
"loss": 0.6016,
|
| 1927 |
+
"step": 27400
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 187.7133105802048,
|
| 1931 |
+
"grad_norm": 0.38929620385169983,
|
| 1932 |
+
"learning_rate": 0.00010849474457892682,
|
| 1933 |
+
"loss": 0.6006,
|
| 1934 |
+
"step": 27500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 188.39590443686006,
|
| 1938 |
+
"grad_norm": 0.42569249868392944,
|
| 1939 |
+
"learning_rate": 0.0001075543026895563,
|
| 1940 |
+
"loss": 0.6008,
|
| 1941 |
+
"step": 27600
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 189.07849829351537,
|
| 1945 |
+
"grad_norm": 0.37430456280708313,
|
| 1946 |
+
"learning_rate": 0.00010666857730420167,
|
| 1947 |
+
"loss": 0.601,
|
| 1948 |
+
"step": 27700
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 189.76109215017064,
|
| 1952 |
+
"grad_norm": 0.4152965545654297,
|
| 1953 |
+
"learning_rate": 0.00010583767795900921,
|
| 1954 |
+
"loss": 0.6006,
|
| 1955 |
+
"step": 27800
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 190.44368600682594,
|
| 1959 |
+
"grad_norm": 0.4082518517971039,
|
| 1960 |
+
"learning_rate": 0.00010506170740988232,
|
| 1961 |
+
"loss": 0.5998,
|
| 1962 |
+
"step": 27900
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 191.12627986348122,
|
| 1966 |
+
"grad_norm": 0.3863820731639862,
|
| 1967 |
+
"learning_rate": 0.00010434076161977406,
|
| 1968 |
+
"loss": 0.6013,
|
| 1969 |
+
"step": 28000
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 191.80887372013652,
|
| 1973 |
+
"grad_norm": 0.36929044127464294,
|
| 1974 |
+
"learning_rate": 0.00010367492974681949,
|
| 1975 |
+
"loss": 0.6007,
|
| 1976 |
+
"step": 28100
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 192.49146757679182,
|
| 1980 |
+
"grad_norm": 0.3993302583694458,
|
| 1981 |
+
"learning_rate": 0.0001030642941333095,
|
| 1982 |
+
"loss": 0.6,
|
| 1983 |
+
"step": 28200
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 193.1740614334471,
|
| 1987 |
+
"grad_norm": 0.36054790019989014,
|
| 1988 |
+
"learning_rate": 0.00010250893029550792,
|
| 1989 |
+
"loss": 0.6006,
|
| 1990 |
+
"step": 28300
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 193.8566552901024,
|
| 1994 |
+
"grad_norm": 0.40515759587287903,
|
| 1995 |
+
"learning_rate": 0.00010200890691431253,
|
| 1996 |
+
"loss": 0.6004,
|
| 1997 |
+
"step": 28400
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 194.53924914675767,
|
| 2001 |
+
"grad_norm": 0.46927398443222046,
|
| 2002 |
+
"learning_rate": 0.00010156428582676107,
|
| 2003 |
+
"loss": 0.5991,
|
| 2004 |
+
"step": 28500
|
| 2005 |
+
}
|
| 2006 |
+
],
|
| 2007 |
+
"logging_steps": 100,
|
| 2008 |
+
"max_steps": 29200,
|
| 2009 |
+
"num_input_tokens_seen": 0,
|
| 2010 |
+
"num_train_epochs": 200,
|
| 2011 |
+
"save_steps": 500,
|
| 2012 |
+
"stateful_callbacks": {
|
| 2013 |
+
"TrainerControl": {
|
| 2014 |
+
"args": {
|
| 2015 |
+
"should_epoch_stop": false,
|
| 2016 |
+
"should_evaluate": false,
|
| 2017 |
+
"should_log": false,
|
| 2018 |
+
"should_save": true,
|
| 2019 |
+
"should_training_stop": false
|
| 2020 |
+
},
|
| 2021 |
+
"attributes": {}
|
| 2022 |
+
}
|
| 2023 |
+
},
|
| 2024 |
+
"total_flos": 9.117195452935045e+17,
|
| 2025 |
+
"train_batch_size": 256,
|
| 2026 |
+
"trial_name": null,
|
| 2027 |
+
"trial_params": null
|
| 2028 |
+
}
|
checkpoint-28500/training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0b3988ed6f3d05a9a0d746b5535dfdb8d9bd5cdf3d3059847445b0b1d44dcf65
|
| 3 |
+
size 5240
|
checkpoint-28500/vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-29000/config.json
ADDED
|
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"activation_function": "gelu_new",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"GPTJForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attn_pdrop": 0.1,
|
| 7 |
+
"bos_token_id": 50256,
|
| 8 |
+
"embd_pdrop": 0.1,
|
| 9 |
+
"eos_token_id": 50256,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"layer_norm_epsilon": 1e-05,
|
| 12 |
+
"model_type": "gptj",
|
| 13 |
+
"n_embd": 768,
|
| 14 |
+
"n_head": 12,
|
| 15 |
+
"n_inner": null,
|
| 16 |
+
"n_layer": 12,
|
| 17 |
+
"n_positions": 1024,
|
| 18 |
+
"resid_pdrop": 0.1,
|
| 19 |
+
"rotary_dim": 64,
|
| 20 |
+
"tie_word_embeddings": false,
|
| 21 |
+
"torch_dtype": "float32",
|
| 22 |
+
"transformers_version": "4.46.2",
|
| 23 |
+
"use_cache": true,
|
| 24 |
+
"vocab_size": 50257
|
| 25 |
+
}
|
checkpoint-29000/generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 50256,
|
| 4 |
+
"eos_token_id": 50256,
|
| 5 |
+
"transformers_version": "4.46.2"
|
| 6 |
+
}
|
checkpoint-29000/merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
checkpoint-29000/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:1741904a4c905b9ed074e9479595d27dd69ba949a11bd42c24932b1b58fd4bbf
|
| 3 |
+
size 648995876
|
checkpoint-29000/optimizer.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ff7292dd97d0109227afcfc2842ae241da3cdcf9c9ee4d1c05eee382883899a7
|
| 3 |
+
size 1298039738
|
checkpoint-29000/rng_state_0.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b56a2c11a60608ad40a120a3fda5535f17e8d6f8730d50d8ab6cac22132ff807
|
| 3 |
+
size 14512
|
checkpoint-29000/rng_state_1.pth
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:31db54fd52f013e8745684dd047cc61d4766fb885923d029ab50eada399bea7b
|
| 3 |
+
size 14512
|
checkpoint-29000/scheduler.pt
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:7c69e56bdf8f3e5565d4850fda4ead6c08749c24392d6b52d0c9cd6d0b9b00bc
|
| 3 |
+
size 1000
|
checkpoint-29000/special_tokens_map.json
ADDED
|
@@ -0,0 +1,24 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<|endoftext|>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": true,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "<|endoftext|>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": true,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": "<|endoftext|>",
|
| 17 |
+
"unk_token": {
|
| 18 |
+
"content": "<|endoftext|>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": true,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
}
|
| 24 |
+
}
|
checkpoint-29000/tokenizer_config.json
ADDED
|
@@ -0,0 +1,22 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"50256": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": true,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
}
|
| 13 |
+
},
|
| 14 |
+
"bos_token": "<|endoftext|>",
|
| 15 |
+
"clean_up_tokenization_spaces": false,
|
| 16 |
+
"eos_token": "<|endoftext|>",
|
| 17 |
+
"errors": "replace",
|
| 18 |
+
"model_max_length": 1024,
|
| 19 |
+
"pad_token": "<|endoftext|>",
|
| 20 |
+
"tokenizer_class": "GPT2Tokenizer",
|
| 21 |
+
"unk_token": "<|endoftext|>"
|
| 22 |
+
}
|
checkpoint-29000/trainer_state.json
ADDED
|
@@ -0,0 +1,2063 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": null,
|
| 3 |
+
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 197.95221843003412,
|
| 5 |
+
"eval_steps": 500,
|
| 6 |
+
"global_step": 29000,
|
| 7 |
+
"is_hyper_param_search": false,
|
| 8 |
+
"is_local_process_zero": true,
|
| 9 |
+
"is_world_process_zero": true,
|
| 10 |
+
"log_history": [
|
| 11 |
+
{
|
| 12 |
+
"epoch": 0.6825938566552902,
|
| 13 |
+
"grad_norm": 1.5864567756652832,
|
| 14 |
+
"learning_rate": 0.0001,
|
| 15 |
+
"loss": 5.292,
|
| 16 |
+
"step": 100
|
| 17 |
+
},
|
| 18 |
+
{
|
| 19 |
+
"epoch": 1.36518771331058,
|
| 20 |
+
"grad_norm": 0.6435245275497437,
|
| 21 |
+
"learning_rate": 0.0002,
|
| 22 |
+
"loss": 1.3885,
|
| 23 |
+
"step": 200
|
| 24 |
+
},
|
| 25 |
+
{
|
| 26 |
+
"epoch": 2.04778156996587,
|
| 27 |
+
"grad_norm": 0.6205620169639587,
|
| 28 |
+
"learning_rate": 0.0003,
|
| 29 |
+
"loss": 1.2326,
|
| 30 |
+
"step": 300
|
| 31 |
+
},
|
| 32 |
+
{
|
| 33 |
+
"epoch": 2.73037542662116,
|
| 34 |
+
"grad_norm": 0.668118953704834,
|
| 35 |
+
"learning_rate": 0.0004,
|
| 36 |
+
"loss": 1.234,
|
| 37 |
+
"step": 400
|
| 38 |
+
},
|
| 39 |
+
{
|
| 40 |
+
"epoch": 3.4129692832764507,
|
| 41 |
+
"grad_norm": 0.5799328684806824,
|
| 42 |
+
"learning_rate": 0.0005,
|
| 43 |
+
"loss": 1.2338,
|
| 44 |
+
"step": 500
|
| 45 |
+
},
|
| 46 |
+
{
|
| 47 |
+
"epoch": 4.09556313993174,
|
| 48 |
+
"grad_norm": 0.5505828261375427,
|
| 49 |
+
"learning_rate": 0.0006,
|
| 50 |
+
"loss": 1.2287,
|
| 51 |
+
"step": 600
|
| 52 |
+
},
|
| 53 |
+
{
|
| 54 |
+
"epoch": 4.778156996587031,
|
| 55 |
+
"grad_norm": 0.515738844871521,
|
| 56 |
+
"learning_rate": 0.0007,
|
| 57 |
+
"loss": 1.2327,
|
| 58 |
+
"step": 700
|
| 59 |
+
},
|
| 60 |
+
{
|
| 61 |
+
"epoch": 5.460750853242321,
|
| 62 |
+
"grad_norm": 0.509695291519165,
|
| 63 |
+
"learning_rate": 0.0008,
|
| 64 |
+
"loss": 1.2565,
|
| 65 |
+
"step": 800
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 6.143344709897611,
|
| 69 |
+
"grad_norm": 0.6063594818115234,
|
| 70 |
+
"learning_rate": 0.0009000000000000001,
|
| 71 |
+
"loss": 1.2289,
|
| 72 |
+
"step": 900
|
| 73 |
+
},
|
| 74 |
+
{
|
| 75 |
+
"epoch": 6.825938566552901,
|
| 76 |
+
"grad_norm": 0.5587486624717712,
|
| 77 |
+
"learning_rate": 0.001,
|
| 78 |
+
"loss": 1.2334,
|
| 79 |
+
"step": 1000
|
| 80 |
+
},
|
| 81 |
+
{
|
| 82 |
+
"epoch": 7.508532423208191,
|
| 83 |
+
"grad_norm": 0.5273250341415405,
|
| 84 |
+
"learning_rate": 0.0009999721746341338,
|
| 85 |
+
"loss": 1.2316,
|
| 86 |
+
"step": 1100
|
| 87 |
+
},
|
| 88 |
+
{
|
| 89 |
+
"epoch": 8.19112627986348,
|
| 90 |
+
"grad_norm": 0.593492329120636,
|
| 91 |
+
"learning_rate": 0.000999888701977651,
|
| 92 |
+
"loss": 1.2279,
|
| 93 |
+
"step": 1200
|
| 94 |
+
},
|
| 95 |
+
{
|
| 96 |
+
"epoch": 8.873720136518772,
|
| 97 |
+
"grad_norm": 0.44249382615089417,
|
| 98 |
+
"learning_rate": 0.000999749592353472,
|
| 99 |
+
"loss": 1.231,
|
| 100 |
+
"step": 1300
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 9.556313993174061,
|
| 104 |
+
"grad_norm": 0.4120616018772125,
|
| 105 |
+
"learning_rate": 0.0009995548629650466,
|
| 106 |
+
"loss": 1.2282,
|
| 107 |
+
"step": 1400
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"epoch": 10.238907849829351,
|
| 111 |
+
"grad_norm": 0.5016219019889832,
|
| 112 |
+
"learning_rate": 0.0009993045378942262,
|
| 113 |
+
"loss": 1.2261,
|
| 114 |
+
"step": 1500
|
| 115 |
+
},
|
| 116 |
+
{
|
| 117 |
+
"epoch": 10.921501706484642,
|
| 118 |
+
"grad_norm": 0.5626952648162842,
|
| 119 |
+
"learning_rate": 0.0009989986480982847,
|
| 120 |
+
"loss": 1.2294,
|
| 121 |
+
"step": 1600
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 11.604095563139932,
|
| 125 |
+
"grad_norm": 0.45088785886764526,
|
| 126 |
+
"learning_rate": 0.000998637231406091,
|
| 127 |
+
"loss": 1.2268,
|
| 128 |
+
"step": 1700
|
| 129 |
+
},
|
| 130 |
+
{
|
| 131 |
+
"epoch": 12.286689419795222,
|
| 132 |
+
"grad_norm": 0.39686086773872375,
|
| 133 |
+
"learning_rate": 0.0009982203325134304,
|
| 134 |
+
"loss": 1.2254,
|
| 135 |
+
"step": 1800
|
| 136 |
+
},
|
| 137 |
+
{
|
| 138 |
+
"epoch": 12.969283276450511,
|
| 139 |
+
"grad_norm": 0.3904390037059784,
|
| 140 |
+
"learning_rate": 0.0009977480029774771,
|
| 141 |
+
"loss": 1.2274,
|
| 142 |
+
"step": 1900
|
| 143 |
+
},
|
| 144 |
+
{
|
| 145 |
+
"epoch": 13.651877133105803,
|
| 146 |
+
"grad_norm": 0.38391488790512085,
|
| 147 |
+
"learning_rate": 0.0009972203012104186,
|
| 148 |
+
"loss": 1.2247,
|
| 149 |
+
"step": 2000
|
| 150 |
+
},
|
| 151 |
+
{
|
| 152 |
+
"epoch": 14.334470989761092,
|
| 153 |
+
"grad_norm": 0.4345687925815582,
|
| 154 |
+
"learning_rate": 0.0009966372924722312,
|
| 155 |
+
"loss": 1.2252,
|
| 156 |
+
"step": 2100
|
| 157 |
+
},
|
| 158 |
+
{
|
| 159 |
+
"epoch": 15.017064846416382,
|
| 160 |
+
"grad_norm": 0.34943118691444397,
|
| 161 |
+
"learning_rate": 0.0009959990488626104,
|
| 162 |
+
"loss": 1.2266,
|
| 163 |
+
"step": 2200
|
| 164 |
+
},
|
| 165 |
+
{
|
| 166 |
+
"epoch": 15.699658703071673,
|
| 167 |
+
"grad_norm": 0.3420363962650299,
|
| 168 |
+
"learning_rate": 0.0009953056493120537,
|
| 169 |
+
"loss": 1.224,
|
| 170 |
+
"step": 2300
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 16.38225255972696,
|
| 174 |
+
"grad_norm": 0.3876005709171295,
|
| 175 |
+
"learning_rate": 0.0009945571795721,
|
| 176 |
+
"loss": 1.2241,
|
| 177 |
+
"step": 2400
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 17.064846416382252,
|
| 181 |
+
"grad_norm": 0.3726520836353302,
|
| 182 |
+
"learning_rate": 0.0009937537322047237,
|
| 183 |
+
"loss": 1.2258,
|
| 184 |
+
"step": 2500
|
| 185 |
+
},
|
| 186 |
+
{
|
| 187 |
+
"epoch": 17.747440273037544,
|
| 188 |
+
"grad_norm": 0.4071378707885742,
|
| 189 |
+
"learning_rate": 0.0009928954065708895,
|
| 190 |
+
"loss": 1.2235,
|
| 191 |
+
"step": 2600
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 18.43003412969283,
|
| 195 |
+
"grad_norm": 0.443796843290329,
|
| 196 |
+
"learning_rate": 0.0009919823088182628,
|
| 197 |
+
"loss": 1.2346,
|
| 198 |
+
"step": 2700
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 19.112627986348123,
|
| 202 |
+
"grad_norm": 0.3463057279586792,
|
| 203 |
+
"learning_rate": 0.000991014551868084,
|
| 204 |
+
"loss": 1.2292,
|
| 205 |
+
"step": 2800
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 19.795221843003414,
|
| 209 |
+
"grad_norm": 0.31187331676483154,
|
| 210 |
+
"learning_rate": 0.0009899922554012027,
|
| 211 |
+
"loss": 1.2237,
|
| 212 |
+
"step": 2900
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 20.477815699658702,
|
| 216 |
+
"grad_norm": 0.3052064776420593,
|
| 217 |
+
"learning_rate": 0.0009889155458432776,
|
| 218 |
+
"loss": 1.2242,
|
| 219 |
+
"step": 3000
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 21.160409556313994,
|
| 223 |
+
"grad_norm": 0.3126174807548523,
|
| 224 |
+
"learning_rate": 0.000987784556349142,
|
| 225 |
+
"loss": 1.2251,
|
| 226 |
+
"step": 3100
|
| 227 |
+
},
|
| 228 |
+
{
|
| 229 |
+
"epoch": 21.843003412969285,
|
| 230 |
+
"grad_norm": 0.29788631200790405,
|
| 231 |
+
"learning_rate": 0.0009865994267863351,
|
| 232 |
+
"loss": 1.2224,
|
| 233 |
+
"step": 3200
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 22.525597269624573,
|
| 237 |
+
"grad_norm": 0.2966502606868744,
|
| 238 |
+
"learning_rate": 0.000985360303717807,
|
| 239 |
+
"loss": 1.2233,
|
| 240 |
+
"step": 3300
|
| 241 |
+
},
|
| 242 |
+
{
|
| 243 |
+
"epoch": 23.208191126279864,
|
| 244 |
+
"grad_norm": 0.31454625725746155,
|
| 245 |
+
"learning_rate": 0.0009840673403837918,
|
| 246 |
+
"loss": 1.2242,
|
| 247 |
+
"step": 3400
|
| 248 |
+
},
|
| 249 |
+
{
|
| 250 |
+
"epoch": 23.890784982935152,
|
| 251 |
+
"grad_norm": 0.33144181966781616,
|
| 252 |
+
"learning_rate": 0.0009827206966828582,
|
| 253 |
+
"loss": 1.2219,
|
| 254 |
+
"step": 3500
|
| 255 |
+
},
|
| 256 |
+
{
|
| 257 |
+
"epoch": 24.573378839590443,
|
| 258 |
+
"grad_norm": 0.3603832423686981,
|
| 259 |
+
"learning_rate": 0.000981320539152133,
|
| 260 |
+
"loss": 1.2227,
|
| 261 |
+
"step": 3600
|
| 262 |
+
},
|
| 263 |
+
{
|
| 264 |
+
"epoch": 25.255972696245735,
|
| 265 |
+
"grad_norm": 0.28107529878616333,
|
| 266 |
+
"learning_rate": 0.0009798670409467075,
|
| 267 |
+
"loss": 1.2236,
|
| 268 |
+
"step": 3700
|
| 269 |
+
},
|
| 270 |
+
{
|
| 271 |
+
"epoch": 25.938566552901023,
|
| 272 |
+
"grad_norm": 0.3241814076900482,
|
| 273 |
+
"learning_rate": 0.0009783603818182237,
|
| 274 |
+
"loss": 1.2212,
|
| 275 |
+
"step": 3800
|
| 276 |
+
},
|
| 277 |
+
{
|
| 278 |
+
"epoch": 26.621160409556314,
|
| 279 |
+
"grad_norm": 0.35018429160118103,
|
| 280 |
+
"learning_rate": 0.0009768007480926437,
|
| 281 |
+
"loss": 1.2223,
|
| 282 |
+
"step": 3900
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"epoch": 27.303754266211605,
|
| 286 |
+
"grad_norm": 0.2918601632118225,
|
| 287 |
+
"learning_rate": 0.0009751883326472076,
|
| 288 |
+
"loss": 1.2228,
|
| 289 |
+
"step": 4000
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 27.986348122866893,
|
| 293 |
+
"grad_norm": 0.29299041628837585,
|
| 294 |
+
"learning_rate": 0.0009735233348865806,
|
| 295 |
+
"loss": 1.2208,
|
| 296 |
+
"step": 4100
|
| 297 |
+
},
|
| 298 |
+
{
|
| 299 |
+
"epoch": 28.668941979522184,
|
| 300 |
+
"grad_norm": 0.3291133642196655,
|
| 301 |
+
"learning_rate": 0.0009718059607181931,
|
| 302 |
+
"loss": 1.2219,
|
| 303 |
+
"step": 4200
|
| 304 |
+
},
|
| 305 |
+
{
|
| 306 |
+
"epoch": 29.351535836177476,
|
| 307 |
+
"grad_norm": 0.2924804985523224,
|
| 308 |
+
"learning_rate": 0.0009700364225267769,
|
| 309 |
+
"loss": 1.2222,
|
| 310 |
+
"step": 4300
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 30.034129692832764,
|
| 314 |
+
"grad_norm": 0.3081340193748474,
|
| 315 |
+
"learning_rate": 0.0009682149391480986,
|
| 316 |
+
"loss": 1.2201,
|
| 317 |
+
"step": 4400
|
| 318 |
+
},
|
| 319 |
+
{
|
| 320 |
+
"epoch": 30.716723549488055,
|
| 321 |
+
"grad_norm": 0.2960987091064453,
|
| 322 |
+
"learning_rate": 0.0009663417358418981,
|
| 323 |
+
"loss": 1.2213,
|
| 324 |
+
"step": 4500
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"epoch": 31.399317406143346,
|
| 328 |
+
"grad_norm": 0.29421886801719666,
|
| 329 |
+
"learning_rate": 0.0009644170442640303,
|
| 330 |
+
"loss": 1.2201,
|
| 331 |
+
"step": 4600
|
| 332 |
+
},
|
| 333 |
+
{
|
| 334 |
+
"epoch": 32.081911262798634,
|
| 335 |
+
"grad_norm": 0.3381507098674774,
|
| 336 |
+
"learning_rate": 0.0009624411024378166,
|
| 337 |
+
"loss": 1.2186,
|
| 338 |
+
"step": 4700
|
| 339 |
+
},
|
| 340 |
+
{
|
| 341 |
+
"epoch": 32.76450511945392,
|
| 342 |
+
"grad_norm": 0.3033551275730133,
|
| 343 |
+
"learning_rate": 0.0009604141547246092,
|
| 344 |
+
"loss": 1.2199,
|
| 345 |
+
"step": 4800
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 33.44709897610922,
|
| 349 |
+
"grad_norm": 0.3156335949897766,
|
| 350 |
+
"learning_rate": 0.0009583364517935713,
|
| 351 |
+
"loss": 1.2189,
|
| 352 |
+
"step": 4900
|
| 353 |
+
},
|
| 354 |
+
{
|
| 355 |
+
"epoch": 34.129692832764505,
|
| 356 |
+
"grad_norm": 0.33952537178993225,
|
| 357 |
+
"learning_rate": 0.0009562082505906771,
|
| 358 |
+
"loss": 1.2163,
|
| 359 |
+
"step": 5000
|
| 360 |
+
},
|
| 361 |
+
{
|
| 362 |
+
"epoch": 34.81228668941979,
|
| 363 |
+
"grad_norm": 0.2941412925720215,
|
| 364 |
+
"learning_rate": 0.0009540298143069361,
|
| 365 |
+
"loss": 1.2179,
|
| 366 |
+
"step": 5100
|
| 367 |
+
},
|
| 368 |
+
{
|
| 369 |
+
"epoch": 35.49488054607509,
|
| 370 |
+
"grad_norm": 0.28668472170829773,
|
| 371 |
+
"learning_rate": 0.0009518014123458442,
|
| 372 |
+
"loss": 1.2158,
|
| 373 |
+
"step": 5200
|
| 374 |
+
},
|
| 375 |
+
{
|
| 376 |
+
"epoch": 36.177474402730375,
|
| 377 |
+
"grad_norm": 0.2789587080478668,
|
| 378 |
+
"learning_rate": 0.0009495233202900674,
|
| 379 |
+
"loss": 1.213,
|
| 380 |
+
"step": 5300
|
| 381 |
+
},
|
| 382 |
+
{
|
| 383 |
+
"epoch": 36.86006825938566,
|
| 384 |
+
"grad_norm": 0.3137347102165222,
|
| 385 |
+
"learning_rate": 0.0009471958198673614,
|
| 386 |
+
"loss": 1.214,
|
| 387 |
+
"step": 5400
|
| 388 |
+
},
|
| 389 |
+
{
|
| 390 |
+
"epoch": 37.54266211604096,
|
| 391 |
+
"grad_norm": 0.25685933232307434,
|
| 392 |
+
"learning_rate": 0.0009448191989157294,
|
| 393 |
+
"loss": 1.21,
|
| 394 |
+
"step": 5500
|
| 395 |
+
},
|
| 396 |
+
{
|
| 397 |
+
"epoch": 38.225255972696246,
|
| 398 |
+
"grad_norm": 0.3252381980419159,
|
| 399 |
+
"learning_rate": 0.0009423937513478271,
|
| 400 |
+
"loss": 1.2071,
|
| 401 |
+
"step": 5600
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 38.907849829351534,
|
| 405 |
+
"grad_norm": 0.2932053208351135,
|
| 406 |
+
"learning_rate": 0.0009399197771146144,
|
| 407 |
+
"loss": 1.2077,
|
| 408 |
+
"step": 5700
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 39.59044368600683,
|
| 412 |
+
"grad_norm": 0.33668607473373413,
|
| 413 |
+
"learning_rate": 0.0009373975821682607,
|
| 414 |
+
"loss": 1.2019,
|
| 415 |
+
"step": 5800
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 40.27303754266212,
|
| 419 |
+
"grad_norm": 0.31952792406082153,
|
| 420 |
+
"learning_rate": 0.0009348274784243094,
|
| 421 |
+
"loss": 1.1979,
|
| 422 |
+
"step": 5900
|
| 423 |
+
},
|
| 424 |
+
{
|
| 425 |
+
"epoch": 40.955631399317404,
|
| 426 |
+
"grad_norm": 0.3525752127170563,
|
| 427 |
+
"learning_rate": 0.0009322097837231025,
|
| 428 |
+
"loss": 1.1999,
|
| 429 |
+
"step": 6000
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 41.6382252559727,
|
| 433 |
+
"grad_norm": 0.3389936685562134,
|
| 434 |
+
"learning_rate": 0.0009295448217904745,
|
| 435 |
+
"loss": 1.1923,
|
| 436 |
+
"step": 6100
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 42.32081911262799,
|
| 440 |
+
"grad_norm": 0.3114762008190155,
|
| 441 |
+
"learning_rate": 0.0009268329221977179,
|
| 442 |
+
"loss": 1.19,
|
| 443 |
+
"step": 6200
|
| 444 |
+
},
|
| 445 |
+
{
|
| 446 |
+
"epoch": 43.003412969283275,
|
| 447 |
+
"grad_norm": 0.3166259229183197,
|
| 448 |
+
"learning_rate": 0.0009240744203208256,
|
| 449 |
+
"loss": 1.1914,
|
| 450 |
+
"step": 6300
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 43.68600682593857,
|
| 454 |
+
"grad_norm": 0.31299668550491333,
|
| 455 |
+
"learning_rate": 0.0009212696572990148,
|
| 456 |
+
"loss": 1.1843,
|
| 457 |
+
"step": 6400
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 44.36860068259386,
|
| 461 |
+
"grad_norm": 0.36728623509407043,
|
| 462 |
+
"learning_rate": 0.0009184189799925403,
|
| 463 |
+
"loss": 1.1828,
|
| 464 |
+
"step": 6500
|
| 465 |
+
},
|
| 466 |
+
{
|
| 467 |
+
"epoch": 45.051194539249146,
|
| 468 |
+
"grad_norm": 0.25386592745780945,
|
| 469 |
+
"learning_rate": 0.0009155227409397977,
|
| 470 |
+
"loss": 1.1849,
|
| 471 |
+
"step": 6600
|
| 472 |
+
},
|
| 473 |
+
{
|
| 474 |
+
"epoch": 45.73378839590444,
|
| 475 |
+
"grad_norm": 0.3170667290687561,
|
| 476 |
+
"learning_rate": 0.0009125812983137259,
|
| 477 |
+
"loss": 1.1789,
|
| 478 |
+
"step": 6700
|
| 479 |
+
},
|
| 480 |
+
{
|
| 481 |
+
"epoch": 46.41638225255973,
|
| 482 |
+
"grad_norm": 0.3354248106479645,
|
| 483 |
+
"learning_rate": 0.0009095950158775128,
|
| 484 |
+
"loss": 1.1781,
|
| 485 |
+
"step": 6800
|
| 486 |
+
},
|
| 487 |
+
{
|
| 488 |
+
"epoch": 47.098976109215016,
|
| 489 |
+
"grad_norm": 0.2761414647102356,
|
| 490 |
+
"learning_rate": 0.0009065642629396087,
|
| 491 |
+
"loss": 1.1803,
|
| 492 |
+
"step": 6900
|
| 493 |
+
},
|
| 494 |
+
{
|
| 495 |
+
"epoch": 47.781569965870304,
|
| 496 |
+
"grad_norm": 0.28061142563819885,
|
| 497 |
+
"learning_rate": 0.0009034894143080554,
|
| 498 |
+
"loss": 1.175,
|
| 499 |
+
"step": 7000
|
| 500 |
+
},
|
| 501 |
+
{
|
| 502 |
+
"epoch": 48.4641638225256,
|
| 503 |
+
"grad_norm": 0.3548315763473511,
|
| 504 |
+
"learning_rate": 0.0009003708502441333,
|
| 505 |
+
"loss": 1.1743,
|
| 506 |
+
"step": 7100
|
| 507 |
+
},
|
| 508 |
+
{
|
| 509 |
+
"epoch": 49.14675767918089,
|
| 510 |
+
"grad_norm": 0.3077629804611206,
|
| 511 |
+
"learning_rate": 0.0008972089564153364,
|
| 512 |
+
"loss": 1.1762,
|
| 513 |
+
"step": 7200
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 49.829351535836174,
|
| 517 |
+
"grad_norm": 0.3600843846797943,
|
| 518 |
+
"learning_rate": 0.0008940041238476759,
|
| 519 |
+
"loss": 1.1718,
|
| 520 |
+
"step": 7300
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 50.51194539249147,
|
| 524 |
+
"grad_norm": 0.3295997083187103,
|
| 525 |
+
"learning_rate": 0.0008907567488773241,
|
| 526 |
+
"loss": 1.1709,
|
| 527 |
+
"step": 7400
|
| 528 |
+
},
|
| 529 |
+
{
|
| 530 |
+
"epoch": 51.19453924914676,
|
| 531 |
+
"grad_norm": 0.32083791494369507,
|
| 532 |
+
"learning_rate": 0.0008874672331015996,
|
| 533 |
+
"loss": 1.1727,
|
| 534 |
+
"step": 7500
|
| 535 |
+
},
|
| 536 |
+
{
|
| 537 |
+
"epoch": 51.877133105802045,
|
| 538 |
+
"grad_norm": 0.32850903272628784,
|
| 539 |
+
"learning_rate": 0.0008841359833293022,
|
| 540 |
+
"loss": 1.1685,
|
| 541 |
+
"step": 7600
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"epoch": 52.55972696245734,
|
| 545 |
+
"grad_norm": 0.33856746554374695,
|
| 546 |
+
"learning_rate": 0.000880763411530404,
|
| 547 |
+
"loss": 1.168,
|
| 548 |
+
"step": 7700
|
| 549 |
+
},
|
| 550 |
+
{
|
| 551 |
+
"epoch": 53.24232081911263,
|
| 552 |
+
"grad_norm": 0.34316346049308777,
|
| 553 |
+
"learning_rate": 0.0008773499347851014,
|
| 554 |
+
"loss": 1.1687,
|
| 555 |
+
"step": 7800
|
| 556 |
+
},
|
| 557 |
+
{
|
| 558 |
+
"epoch": 53.924914675767916,
|
| 559 |
+
"grad_norm": 0.38703614473342896,
|
| 560 |
+
"learning_rate": 0.0008738959752322358,
|
| 561 |
+
"loss": 1.165,
|
| 562 |
+
"step": 7900
|
| 563 |
+
},
|
| 564 |
+
{
|
| 565 |
+
"epoch": 54.60750853242321,
|
| 566 |
+
"grad_norm": 0.3954058587551117,
|
| 567 |
+
"learning_rate": 0.0008704019600170884,
|
| 568 |
+
"loss": 1.1637,
|
| 569 |
+
"step": 8000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 55.2901023890785,
|
| 573 |
+
"grad_norm": 0.33417293429374695,
|
| 574 |
+
"learning_rate": 0.0008668683212385556,
|
| 575 |
+
"loss": 1.1635,
|
| 576 |
+
"step": 8100
|
| 577 |
+
},
|
| 578 |
+
{
|
| 579 |
+
"epoch": 55.972696245733786,
|
| 580 |
+
"grad_norm": 0.3305743932723999,
|
| 581 |
+
"learning_rate": 0.0008632954958957123,
|
| 582 |
+
"loss": 1.1601,
|
| 583 |
+
"step": 8200
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"epoch": 56.65529010238908,
|
| 587 |
+
"grad_norm": 0.3755279779434204,
|
| 588 |
+
"learning_rate": 0.0008596839258337689,
|
| 589 |
+
"loss": 1.158,
|
| 590 |
+
"step": 8300
|
| 591 |
+
},
|
| 592 |
+
{
|
| 593 |
+
"epoch": 57.33788395904437,
|
| 594 |
+
"grad_norm": 0.3553084433078766,
|
| 595 |
+
"learning_rate": 0.000856034057689429,
|
| 596 |
+
"loss": 1.1568,
|
| 597 |
+
"step": 8400
|
| 598 |
+
},
|
| 599 |
+
{
|
| 600 |
+
"epoch": 58.02047781569966,
|
| 601 |
+
"grad_norm": 0.38361021876335144,
|
| 602 |
+
"learning_rate": 0.0008523463428356549,
|
| 603 |
+
"loss": 1.154,
|
| 604 |
+
"step": 8500
|
| 605 |
+
},
|
| 606 |
+
{
|
| 607 |
+
"epoch": 58.70307167235495,
|
| 608 |
+
"grad_norm": 0.35759079456329346,
|
| 609 |
+
"learning_rate": 0.0008486212373258465,
|
| 610 |
+
"loss": 1.152,
|
| 611 |
+
"step": 8600
|
| 612 |
+
},
|
| 613 |
+
{
|
| 614 |
+
"epoch": 59.38566552901024,
|
| 615 |
+
"grad_norm": 0.32452645897865295,
|
| 616 |
+
"learning_rate": 0.0008448592018374424,
|
| 617 |
+
"loss": 1.1497,
|
| 618 |
+
"step": 8700
|
| 619 |
+
},
|
| 620 |
+
{
|
| 621 |
+
"epoch": 60.06825938566553,
|
| 622 |
+
"grad_norm": 0.37652239203453064,
|
| 623 |
+
"learning_rate": 0.0008410607016149491,
|
| 624 |
+
"loss": 1.1461,
|
| 625 |
+
"step": 8800
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 60.75085324232082,
|
| 629 |
+
"grad_norm": 0.35698965191841125,
|
| 630 |
+
"learning_rate": 0.0008372262064124035,
|
| 631 |
+
"loss": 1.1453,
|
| 632 |
+
"step": 8900
|
| 633 |
+
},
|
| 634 |
+
{
|
| 635 |
+
"epoch": 61.43344709897611,
|
| 636 |
+
"grad_norm": 0.3505449891090393,
|
| 637 |
+
"learning_rate": 0.0008333561904352813,
|
| 638 |
+
"loss": 1.1409,
|
| 639 |
+
"step": 9000
|
| 640 |
+
},
|
| 641 |
+
{
|
| 642 |
+
"epoch": 62.1160409556314,
|
| 643 |
+
"grad_norm": 0.3646785616874695,
|
| 644 |
+
"learning_rate": 0.000829451132281851,
|
| 645 |
+
"loss": 1.1363,
|
| 646 |
+
"step": 9100
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 62.79863481228669,
|
| 650 |
+
"grad_norm": 0.34945082664489746,
|
| 651 |
+
"learning_rate": 0.0008255115148839877,
|
| 652 |
+
"loss": 1.1371,
|
| 653 |
+
"step": 9200
|
| 654 |
+
},
|
| 655 |
+
{
|
| 656 |
+
"epoch": 63.48122866894198,
|
| 657 |
+
"grad_norm": 0.34238216280937195,
|
| 658 |
+
"learning_rate": 0.0008215378254474485,
|
| 659 |
+
"loss": 1.1296,
|
| 660 |
+
"step": 9300
|
| 661 |
+
},
|
| 662 |
+
{
|
| 663 |
+
"epoch": 64.16382252559727,
|
| 664 |
+
"grad_norm": 0.37401944398880005,
|
| 665 |
+
"learning_rate": 0.0008175305553916226,
|
| 666 |
+
"loss": 1.1241,
|
| 667 |
+
"step": 9400
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 64.84641638225256,
|
| 671 |
+
"grad_norm": 0.3763269782066345,
|
| 672 |
+
"learning_rate": 0.000813490200288756,
|
| 673 |
+
"loss": 1.1257,
|
| 674 |
+
"step": 9500
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 65.52901023890784,
|
| 678 |
+
"grad_norm": 0.3553602993488312,
|
| 679 |
+
"learning_rate": 0.0008094172598026662,
|
| 680 |
+
"loss": 1.1147,
|
| 681 |
+
"step": 9600
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 66.21160409556315,
|
| 685 |
+
"grad_norm": 0.36700302362442017,
|
| 686 |
+
"learning_rate": 0.0008053122376269497,
|
| 687 |
+
"loss": 1.1078,
|
| 688 |
+
"step": 9700
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 66.89419795221843,
|
| 692 |
+
"grad_norm": 0.4266979694366455,
|
| 693 |
+
"learning_rate": 0.0008011756414226908,
|
| 694 |
+
"loss": 1.1114,
|
| 695 |
+
"step": 9800
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 67.57679180887372,
|
| 699 |
+
"grad_norm": 0.3603265583515167,
|
| 700 |
+
"learning_rate": 0.0007970079827556795,
|
| 701 |
+
"loss": 1.0967,
|
| 702 |
+
"step": 9900
|
| 703 |
+
},
|
| 704 |
+
{
|
| 705 |
+
"epoch": 68.25938566552901,
|
| 706 |
+
"grad_norm": 0.39329585433006287,
|
| 707 |
+
"learning_rate": 0.000792809777033148,
|
| 708 |
+
"loss": 1.0898,
|
| 709 |
+
"step": 10000
|
| 710 |
+
},
|
| 711 |
+
{
|
| 712 |
+
"epoch": 68.9419795221843,
|
| 713 |
+
"grad_norm": 0.40609198808670044,
|
| 714 |
+
"learning_rate": 0.0007885815434400306,
|
| 715 |
+
"loss": 1.0944,
|
| 716 |
+
"step": 10100
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 69.62457337883959,
|
| 720 |
+
"grad_norm": 0.3622093200683594,
|
| 721 |
+
"learning_rate": 0.0007843238048747572,
|
| 722 |
+
"loss": 1.0761,
|
| 723 |
+
"step": 10200
|
| 724 |
+
},
|
| 725 |
+
{
|
| 726 |
+
"epoch": 70.30716723549489,
|
| 727 |
+
"grad_norm": 0.39394432306289673,
|
| 728 |
+
"learning_rate": 0.0007800370878845866,
|
| 729 |
+
"loss": 1.0692,
|
| 730 |
+
"step": 10300
|
| 731 |
+
},
|
| 732 |
+
{
|
| 733 |
+
"epoch": 70.98976109215018,
|
| 734 |
+
"grad_norm": 0.382322758436203,
|
| 735 |
+
"learning_rate": 0.0007757219226004912,
|
| 736 |
+
"loss": 1.0767,
|
| 737 |
+
"step": 10400
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 71.67235494880546,
|
| 741 |
+
"grad_norm": 0.3804159462451935,
|
| 742 |
+
"learning_rate": 0.0007713788426715937,
|
| 743 |
+
"loss": 1.0543,
|
| 744 |
+
"step": 10500
|
| 745 |
+
},
|
| 746 |
+
{
|
| 747 |
+
"epoch": 72.35494880546075,
|
| 748 |
+
"grad_norm": 0.3960065245628357,
|
| 749 |
+
"learning_rate": 0.0007670083851991749,
|
| 750 |
+
"loss": 1.0477,
|
| 751 |
+
"step": 10600
|
| 752 |
+
},
|
| 753 |
+
{
|
| 754 |
+
"epoch": 73.03754266211604,
|
| 755 |
+
"grad_norm": 0.39962372183799744,
|
| 756 |
+
"learning_rate": 0.0007626110906702484,
|
| 757 |
+
"loss": 1.0561,
|
| 758 |
+
"step": 10700
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"epoch": 73.72013651877133,
|
| 762 |
+
"grad_norm": 0.4019100069999695,
|
| 763 |
+
"learning_rate": 0.0007581875028907215,
|
| 764 |
+
"loss": 1.0329,
|
| 765 |
+
"step": 10800
|
| 766 |
+
},
|
| 767 |
+
{
|
| 768 |
+
"epoch": 74.40273037542661,
|
| 769 |
+
"grad_norm": 0.42911744117736816,
|
| 770 |
+
"learning_rate": 0.0007537381689181425,
|
| 771 |
+
"loss": 1.0265,
|
| 772 |
+
"step": 10900
|
| 773 |
+
},
|
| 774 |
+
{
|
| 775 |
+
"epoch": 75.08532423208192,
|
| 776 |
+
"grad_norm": 0.4002211093902588,
|
| 777 |
+
"learning_rate": 0.0007492636389940478,
|
| 778 |
+
"loss": 1.0332,
|
| 779 |
+
"step": 11000
|
| 780 |
+
},
|
| 781 |
+
{
|
| 782 |
+
"epoch": 75.7679180887372,
|
| 783 |
+
"grad_norm": 0.40011659264564514,
|
| 784 |
+
"learning_rate": 0.0007447644664759138,
|
| 785 |
+
"loss": 1.0127,
|
| 786 |
+
"step": 11100
|
| 787 |
+
},
|
| 788 |
+
{
|
| 789 |
+
"epoch": 76.45051194539249,
|
| 790 |
+
"grad_norm": 0.43485236167907715,
|
| 791 |
+
"learning_rate": 0.0007402412077687246,
|
| 792 |
+
"loss": 1.0047,
|
| 793 |
+
"step": 11200
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 77.13310580204778,
|
| 797 |
+
"grad_norm": 0.4095554053783417,
|
| 798 |
+
"learning_rate": 0.0007356944222561622,
|
| 799 |
+
"loss": 1.0114,
|
| 800 |
+
"step": 11300
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"epoch": 77.81569965870307,
|
| 804 |
+
"grad_norm": 0.39631539583206177,
|
| 805 |
+
"learning_rate": 0.0007311246722314281,
|
| 806 |
+
"loss": 0.9916,
|
| 807 |
+
"step": 11400
|
| 808 |
+
},
|
| 809 |
+
{
|
| 810 |
+
"epoch": 78.49829351535836,
|
| 811 |
+
"grad_norm": 0.4598711133003235,
|
| 812 |
+
"learning_rate": 0.0007265325228277065,
|
| 813 |
+
"loss": 0.9822,
|
| 814 |
+
"step": 11500
|
| 815 |
+
},
|
| 816 |
+
{
|
| 817 |
+
"epoch": 79.18088737201366,
|
| 818 |
+
"grad_norm": 0.39505425095558167,
|
| 819 |
+
"learning_rate": 0.0007219185419482739,
|
| 820 |
+
"loss": 0.9883,
|
| 821 |
+
"step": 11600
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 79.86348122866895,
|
| 825 |
+
"grad_norm": 0.3844675123691559,
|
| 826 |
+
"learning_rate": 0.0007172833001962683,
|
| 827 |
+
"loss": 0.9701,
|
| 828 |
+
"step": 11700
|
| 829 |
+
},
|
| 830 |
+
{
|
| 831 |
+
"epoch": 80.54607508532423,
|
| 832 |
+
"grad_norm": 0.48971033096313477,
|
| 833 |
+
"learning_rate": 0.000712627370804124,
|
| 834 |
+
"loss": 0.9597,
|
| 835 |
+
"step": 11800
|
| 836 |
+
},
|
| 837 |
+
{
|
| 838 |
+
"epoch": 81.22866894197952,
|
| 839 |
+
"grad_norm": 0.40114977955818176,
|
| 840 |
+
"learning_rate": 0.0007079513295626796,
|
| 841 |
+
"loss": 0.9643,
|
| 842 |
+
"step": 11900
|
| 843 |
+
},
|
| 844 |
+
{
|
| 845 |
+
"epoch": 81.91126279863481,
|
| 846 |
+
"grad_norm": 0.42664656043052673,
|
| 847 |
+
"learning_rate": 0.0007032557547499727,
|
| 848 |
+
"loss": 0.9482,
|
| 849 |
+
"step": 12000
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 82.5938566552901,
|
| 853 |
+
"grad_norm": 0.5133378505706787,
|
| 854 |
+
"learning_rate": 0.0006985412270597231,
|
| 855 |
+
"loss": 0.9363,
|
| 856 |
+
"step": 12100
|
| 857 |
+
},
|
| 858 |
+
{
|
| 859 |
+
"epoch": 83.2764505119454,
|
| 860 |
+
"grad_norm": 0.43057283759117126,
|
| 861 |
+
"learning_rate": 0.000693808329529522,
|
| 862 |
+
"loss": 0.9409,
|
| 863 |
+
"step": 12200
|
| 864 |
+
},
|
| 865 |
+
{
|
| 866 |
+
"epoch": 83.95904436860069,
|
| 867 |
+
"grad_norm": 0.4190930426120758,
|
| 868 |
+
"learning_rate": 0.0006890576474687264,
|
| 869 |
+
"loss": 0.9255,
|
| 870 |
+
"step": 12300
|
| 871 |
+
},
|
| 872 |
+
{
|
| 873 |
+
"epoch": 84.64163822525597,
|
| 874 |
+
"grad_norm": 0.43727907538414,
|
| 875 |
+
"learning_rate": 0.000684289768386076,
|
| 876 |
+
"loss": 0.9119,
|
| 877 |
+
"step": 12400
|
| 878 |
+
},
|
| 879 |
+
{
|
| 880 |
+
"epoch": 85.32423208191126,
|
| 881 |
+
"grad_norm": 0.4994739890098572,
|
| 882 |
+
"learning_rate": 0.000679505281917037,
|
| 883 |
+
"loss": 0.914,
|
| 884 |
+
"step": 12500
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 86.00682593856655,
|
| 888 |
+
"grad_norm": 0.45477139949798584,
|
| 889 |
+
"learning_rate": 0.0006747047797508828,
|
| 890 |
+
"loss": 0.9017,
|
| 891 |
+
"step": 12600
|
| 892 |
+
},
|
| 893 |
+
{
|
| 894 |
+
"epoch": 86.68941979522184,
|
| 895 |
+
"grad_norm": 0.4953708052635193,
|
| 896 |
+
"learning_rate": 0.0006698888555575204,
|
| 897 |
+
"loss": 0.8881,
|
| 898 |
+
"step": 12700
|
| 899 |
+
},
|
| 900 |
+
{
|
| 901 |
+
"epoch": 87.37201365187714,
|
| 902 |
+
"grad_norm": 0.4449358284473419,
|
| 903 |
+
"learning_rate": 0.000665058104914073,
|
| 904 |
+
"loss": 0.8875,
|
| 905 |
+
"step": 12800
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 88.05460750853243,
|
| 909 |
+
"grad_norm": 0.4819263815879822,
|
| 910 |
+
"learning_rate": 0.0006602131252312249,
|
| 911 |
+
"loss": 0.8748,
|
| 912 |
+
"step": 12900
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 88.73720136518772,
|
| 916 |
+
"grad_norm": 0.5412826538085938,
|
| 917 |
+
"learning_rate": 0.0006553545156793422,
|
| 918 |
+
"loss": 0.8655,
|
| 919 |
+
"step": 13000
|
| 920 |
+
},
|
| 921 |
+
{
|
| 922 |
+
"epoch": 89.419795221843,
|
| 923 |
+
"grad_norm": 0.4753580391407013,
|
| 924 |
+
"learning_rate": 0.0006504828771143733,
|
| 925 |
+
"loss": 0.8617,
|
| 926 |
+
"step": 13100
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 90.10238907849829,
|
| 930 |
+
"grad_norm": 0.545566737651825,
|
| 931 |
+
"learning_rate": 0.0006455988120035418,
|
| 932 |
+
"loss": 0.85,
|
| 933 |
+
"step": 13200
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 90.78498293515358,
|
| 937 |
+
"grad_norm": 0.5255779027938843,
|
| 938 |
+
"learning_rate": 0.0006407029243508423,
|
| 939 |
+
"loss": 0.8449,
|
| 940 |
+
"step": 13300
|
| 941 |
+
},
|
| 942 |
+
{
|
| 943 |
+
"epoch": 91.46757679180887,
|
| 944 |
+
"grad_norm": 0.43862029910087585,
|
| 945 |
+
"learning_rate": 0.0006357958196223419,
|
| 946 |
+
"loss": 0.8371,
|
| 947 |
+
"step": 13400
|
| 948 |
+
},
|
| 949 |
+
{
|
| 950 |
+
"epoch": 92.15017064846417,
|
| 951 |
+
"grad_norm": 0.5360426902770996,
|
| 952 |
+
"learning_rate": 0.0006308781046713046,
|
| 953 |
+
"loss": 0.8257,
|
| 954 |
+
"step": 13500
|
| 955 |
+
},
|
| 956 |
+
{
|
| 957 |
+
"epoch": 92.83276450511946,
|
| 958 |
+
"grad_norm": 0.5451890230178833,
|
| 959 |
+
"learning_rate": 0.0006259503876631428,
|
| 960 |
+
"loss": 0.8247,
|
| 961 |
+
"step": 13600
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 93.51535836177474,
|
| 965 |
+
"grad_norm": 0.4524667263031006,
|
| 966 |
+
"learning_rate": 0.0006210132780002057,
|
| 967 |
+
"loss": 0.8147,
|
| 968 |
+
"step": 13700
|
| 969 |
+
},
|
| 970 |
+
{
|
| 971 |
+
"epoch": 94.19795221843003,
|
| 972 |
+
"grad_norm": 0.6203665733337402,
|
| 973 |
+
"learning_rate": 0.0006160673862464167,
|
| 974 |
+
"loss": 0.8022,
|
| 975 |
+
"step": 13800
|
| 976 |
+
},
|
| 977 |
+
{
|
| 978 |
+
"epoch": 94.88054607508532,
|
| 979 |
+
"grad_norm": 0.5576283931732178,
|
| 980 |
+
"learning_rate": 0.0006111133240517648,
|
| 981 |
+
"loss": 0.8057,
|
| 982 |
+
"step": 13900
|
| 983 |
+
},
|
| 984 |
+
{
|
| 985 |
+
"epoch": 95.56313993174061,
|
| 986 |
+
"grad_norm": 0.48380613327026367,
|
| 987 |
+
"learning_rate": 0.000606151704076664,
|
| 988 |
+
"loss": 0.7906,
|
| 989 |
+
"step": 14000
|
| 990 |
+
},
|
| 991 |
+
{
|
| 992 |
+
"epoch": 96.24573378839591,
|
| 993 |
+
"grad_norm": 0.5328605771064758,
|
| 994 |
+
"learning_rate": 0.0006011831399161856,
|
| 995 |
+
"loss": 0.7823,
|
| 996 |
+
"step": 14100
|
| 997 |
+
},
|
| 998 |
+
{
|
| 999 |
+
"epoch": 96.9283276450512,
|
| 1000 |
+
"grad_norm": 0.5395786762237549,
|
| 1001 |
+
"learning_rate": 0.0005962082460241767,
|
| 1002 |
+
"loss": 0.7883,
|
| 1003 |
+
"step": 14200
|
| 1004 |
+
},
|
| 1005 |
+
{
|
| 1006 |
+
"epoch": 97.61092150170649,
|
| 1007 |
+
"grad_norm": 0.48329558968544006,
|
| 1008 |
+
"learning_rate": 0.0005912276376372715,
|
| 1009 |
+
"loss": 0.7707,
|
| 1010 |
+
"step": 14300
|
| 1011 |
+
},
|
| 1012 |
+
{
|
| 1013 |
+
"epoch": 98.29351535836177,
|
| 1014 |
+
"grad_norm": 0.5496215224266052,
|
| 1015 |
+
"learning_rate": 0.0005862419306988061,
|
| 1016 |
+
"loss": 0.7636,
|
| 1017 |
+
"step": 14400
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 98.97610921501706,
|
| 1021 |
+
"grad_norm": 0.6007915139198303,
|
| 1022 |
+
"learning_rate": 0.0005812517417826456,
|
| 1023 |
+
"loss": 0.7712,
|
| 1024 |
+
"step": 14500
|
| 1025 |
+
},
|
| 1026 |
+
{
|
| 1027 |
+
"epoch": 99.65870307167235,
|
| 1028 |
+
"grad_norm": 0.4981062710285187,
|
| 1029 |
+
"learning_rate": 0.000576257688016934,
|
| 1030 |
+
"loss": 0.7529,
|
| 1031 |
+
"step": 14600
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 100.34129692832765,
|
| 1035 |
+
"grad_norm": 0.5475151538848877,
|
| 1036 |
+
"learning_rate": 0.0005712603870077736,
|
| 1037 |
+
"loss": 0.7468,
|
| 1038 |
+
"step": 14700
|
| 1039 |
+
},
|
| 1040 |
+
{
|
| 1041 |
+
"epoch": 101.02389078498294,
|
| 1042 |
+
"grad_norm": 0.4879900813102722,
|
| 1043 |
+
"learning_rate": 0.0005662604567628494,
|
| 1044 |
+
"loss": 0.7568,
|
| 1045 |
+
"step": 14800
|
| 1046 |
+
},
|
| 1047 |
+
{
|
| 1048 |
+
"epoch": 101.70648464163823,
|
| 1049 |
+
"grad_norm": 0.4963570535182953,
|
| 1050 |
+
"learning_rate": 0.0005612585156149982,
|
| 1051 |
+
"loss": 0.7386,
|
| 1052 |
+
"step": 14900
|
| 1053 |
+
},
|
| 1054 |
+
{
|
| 1055 |
+
"epoch": 102.38907849829351,
|
| 1056 |
+
"grad_norm": 0.5414558053016663,
|
| 1057 |
+
"learning_rate": 0.0005562551821457418,
|
| 1058 |
+
"loss": 0.7326,
|
| 1059 |
+
"step": 15000
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"epoch": 103.0716723549488,
|
| 1063 |
+
"grad_norm": 0.47427278757095337,
|
| 1064 |
+
"learning_rate": 0.0005512510751087889,
|
| 1065 |
+
"loss": 0.7415,
|
| 1066 |
+
"step": 15100
|
| 1067 |
+
},
|
| 1068 |
+
{
|
| 1069 |
+
"epoch": 103.75426621160409,
|
| 1070 |
+
"grad_norm": 0.48589903116226196,
|
| 1071 |
+
"learning_rate": 0.0005462468133535131,
|
| 1072 |
+
"loss": 0.7267,
|
| 1073 |
+
"step": 15200
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 104.43686006825939,
|
| 1077 |
+
"grad_norm": 0.5541400909423828,
|
| 1078 |
+
"learning_rate": 0.0005412430157484224,
|
| 1079 |
+
"loss": 0.7206,
|
| 1080 |
+
"step": 15300
|
| 1081 |
+
},
|
| 1082 |
+
{
|
| 1083 |
+
"epoch": 105.11945392491468,
|
| 1084 |
+
"grad_norm": 0.4941166341304779,
|
| 1085 |
+
"learning_rate": 0.0005362403011046234,
|
| 1086 |
+
"loss": 0.7289,
|
| 1087 |
+
"step": 15400
|
| 1088 |
+
},
|
| 1089 |
+
{
|
| 1090 |
+
"epoch": 105.80204778156997,
|
| 1091 |
+
"grad_norm": 0.4855073094367981,
|
| 1092 |
+
"learning_rate": 0.0005312392880992954,
|
| 1093 |
+
"loss": 0.716,
|
| 1094 |
+
"step": 15500
|
| 1095 |
+
},
|
| 1096 |
+
{
|
| 1097 |
+
"epoch": 106.48464163822526,
|
| 1098 |
+
"grad_norm": 0.5357630252838135,
|
| 1099 |
+
"learning_rate": 0.0005262405951991788,
|
| 1100 |
+
"loss": 0.7111,
|
| 1101 |
+
"step": 15600
|
| 1102 |
+
},
|
| 1103 |
+
{
|
| 1104 |
+
"epoch": 107.16723549488054,
|
| 1105 |
+
"grad_norm": 0.5147479772567749,
|
| 1106 |
+
"learning_rate": 0.0005212448405840905,
|
| 1107 |
+
"loss": 0.7181,
|
| 1108 |
+
"step": 15700
|
| 1109 |
+
},
|
| 1110 |
+
{
|
| 1111 |
+
"epoch": 107.84982935153583,
|
| 1112 |
+
"grad_norm": 0.47526028752326965,
|
| 1113 |
+
"learning_rate": 0.0005162526420704744,
|
| 1114 |
+
"loss": 0.7072,
|
| 1115 |
+
"step": 15800
|
| 1116 |
+
},
|
| 1117 |
+
{
|
| 1118 |
+
"epoch": 108.53242320819113,
|
| 1119 |
+
"grad_norm": 0.528582751750946,
|
| 1120 |
+
"learning_rate": 0.0005112646170349981,
|
| 1121 |
+
"loss": 0.7021,
|
| 1122 |
+
"step": 15900
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 109.21501706484642,
|
| 1126 |
+
"grad_norm": 0.4804699420928955,
|
| 1127 |
+
"learning_rate": 0.0005062813823382014,
|
| 1128 |
+
"loss": 0.7093,
|
| 1129 |
+
"step": 16000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 109.89761092150171,
|
| 1133 |
+
"grad_norm": 0.5126569271087646,
|
| 1134 |
+
"learning_rate": 0.000501303554248212,
|
| 1135 |
+
"loss": 0.7008,
|
| 1136 |
+
"step": 16100
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 110.580204778157,
|
| 1140 |
+
"grad_norm": 0.5352160930633545,
|
| 1141 |
+
"learning_rate": 0.0004963317483645311,
|
| 1142 |
+
"loss": 0.6946,
|
| 1143 |
+
"step": 16200
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 111.26279863481228,
|
| 1147 |
+
"grad_norm": 0.5465342402458191,
|
| 1148 |
+
"learning_rate": 0.0004913665795419041,
|
| 1149 |
+
"loss": 0.7008,
|
| 1150 |
+
"step": 16300
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 111.94539249146757,
|
| 1154 |
+
"grad_norm": 0.45367056131362915,
|
| 1155 |
+
"learning_rate": 0.00048640866181428265,
|
| 1156 |
+
"loss": 0.6939,
|
| 1157 |
+
"step": 16400
|
| 1158 |
+
},
|
| 1159 |
+
{
|
| 1160 |
+
"epoch": 112.62798634812286,
|
| 1161 |
+
"grad_norm": 0.6308451294898987,
|
| 1162 |
+
"learning_rate": 0.00048145860831888776,
|
| 1163 |
+
"loss": 0.6885,
|
| 1164 |
+
"step": 16500
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 113.31058020477816,
|
| 1168 |
+
"grad_norm": 0.4779800772666931,
|
| 1169 |
+
"learning_rate": 0.00047651703122038395,
|
| 1170 |
+
"loss": 0.6939,
|
| 1171 |
+
"step": 16600
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 113.99317406143345,
|
| 1175 |
+
"grad_norm": 0.4980040192604065,
|
| 1176 |
+
"learning_rate": 0.0004715845416351744,
|
| 1177 |
+
"loss": 0.6879,
|
| 1178 |
+
"step": 16700
|
| 1179 |
+
},
|
| 1180 |
+
{
|
| 1181 |
+
"epoch": 114.67576791808874,
|
| 1182 |
+
"grad_norm": 0.5260293483734131,
|
| 1183 |
+
"learning_rate": 0.00046666174955582507,
|
| 1184 |
+
"loss": 0.6835,
|
| 1185 |
+
"step": 16800
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 115.35836177474403,
|
| 1189 |
+
"grad_norm": 0.47831013798713684,
|
| 1190 |
+
"learning_rate": 0.0004617492637756282,
|
| 1191 |
+
"loss": 0.6877,
|
| 1192 |
+
"step": 16900
|
| 1193 |
+
},
|
| 1194 |
+
{
|
| 1195 |
+
"epoch": 116.04095563139931,
|
| 1196 |
+
"grad_norm": 0.5765718221664429,
|
| 1197 |
+
"learning_rate": 0.0004568476918133129,
|
| 1198 |
+
"loss": 0.6813,
|
| 1199 |
+
"step": 17000
|
| 1200 |
+
},
|
| 1201 |
+
{
|
| 1202 |
+
"epoch": 116.7235494880546,
|
| 1203 |
+
"grad_norm": 0.5322794318199158,
|
| 1204 |
+
"learning_rate": 0.00045195763983791573,
|
| 1205 |
+
"loss": 0.679,
|
| 1206 |
+
"step": 17100
|
| 1207 |
+
},
|
| 1208 |
+
{
|
| 1209 |
+
"epoch": 117.4061433447099,
|
| 1210 |
+
"grad_norm": 0.514970064163208,
|
| 1211 |
+
"learning_rate": 0.00044707971259381584,
|
| 1212 |
+
"loss": 0.68,
|
| 1213 |
+
"step": 17200
|
| 1214 |
+
},
|
| 1215 |
+
{
|
| 1216 |
+
"epoch": 118.08873720136519,
|
| 1217 |
+
"grad_norm": 0.539171576499939,
|
| 1218 |
+
"learning_rate": 0.00044221451332594785,
|
| 1219 |
+
"loss": 0.6769,
|
| 1220 |
+
"step": 17300
|
| 1221 |
+
},
|
| 1222 |
+
{
|
| 1223 |
+
"epoch": 118.77133105802048,
|
| 1224 |
+
"grad_norm": 0.5568128824234009,
|
| 1225 |
+
"learning_rate": 0.00043736264370519863,
|
| 1226 |
+
"loss": 0.6757,
|
| 1227 |
+
"step": 17400
|
| 1228 |
+
},
|
| 1229 |
+
{
|
| 1230 |
+
"epoch": 119.45392491467577,
|
| 1231 |
+
"grad_norm": 0.4861965477466583,
|
| 1232 |
+
"learning_rate": 0.0004325247037540011,
|
| 1233 |
+
"loss": 0.6749,
|
| 1234 |
+
"step": 17500
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"epoch": 120.13651877133105,
|
| 1238 |
+
"grad_norm": 0.49631592631340027,
|
| 1239 |
+
"learning_rate": 0.00042770129177212914,
|
| 1240 |
+
"loss": 0.671,
|
| 1241 |
+
"step": 17600
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 120.81911262798634,
|
| 1245 |
+
"grad_norm": 0.5209415555000305,
|
| 1246 |
+
"learning_rate": 0.00042289300426270806,
|
| 1247 |
+
"loss": 0.6714,
|
| 1248 |
+
"step": 17700
|
| 1249 |
+
},
|
| 1250 |
+
{
|
| 1251 |
+
"epoch": 121.50170648464164,
|
| 1252 |
+
"grad_norm": 0.4455106556415558,
|
| 1253 |
+
"learning_rate": 0.0004181004358584447,
|
| 1254 |
+
"loss": 0.6699,
|
| 1255 |
+
"step": 17800
|
| 1256 |
+
},
|
| 1257 |
+
{
|
| 1258 |
+
"epoch": 122.18430034129693,
|
| 1259 |
+
"grad_norm": 0.46950340270996094,
|
| 1260 |
+
"learning_rate": 0.00041332417924809175,
|
| 1261 |
+
"loss": 0.6654,
|
| 1262 |
+
"step": 17900
|
| 1263 |
+
},
|
| 1264 |
+
{
|
| 1265 |
+
"epoch": 122.86689419795222,
|
| 1266 |
+
"grad_norm": 0.5346481800079346,
|
| 1267 |
+
"learning_rate": 0.00040856482510314985,
|
| 1268 |
+
"loss": 0.6681,
|
| 1269 |
+
"step": 18000
|
| 1270 |
+
},
|
| 1271 |
+
{
|
| 1272 |
+
"epoch": 123.54948805460751,
|
| 1273 |
+
"grad_norm": 0.47252893447875977,
|
| 1274 |
+
"learning_rate": 0.0004038229620048207,
|
| 1275 |
+
"loss": 0.6648,
|
| 1276 |
+
"step": 18100
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"epoch": 124.2320819112628,
|
| 1280 |
+
"grad_norm": 0.5372588634490967,
|
| 1281 |
+
"learning_rate": 0.00039909917637121796,
|
| 1282 |
+
"loss": 0.6611,
|
| 1283 |
+
"step": 18200
|
| 1284 |
+
},
|
| 1285 |
+
{
|
| 1286 |
+
"epoch": 124.91467576791808,
|
| 1287 |
+
"grad_norm": 0.5461378693580627,
|
| 1288 |
+
"learning_rate": 0.00039439405238484614,
|
| 1289 |
+
"loss": 0.6637,
|
| 1290 |
+
"step": 18300
|
| 1291 |
+
},
|
| 1292 |
+
{
|
| 1293 |
+
"epoch": 125.59726962457339,
|
| 1294 |
+
"grad_norm": 0.4744736850261688,
|
| 1295 |
+
"learning_rate": 0.0003897081719203558,
|
| 1296 |
+
"loss": 0.659,
|
| 1297 |
+
"step": 18400
|
| 1298 |
+
},
|
| 1299 |
+
{
|
| 1300 |
+
"epoch": 126.27986348122867,
|
| 1301 |
+
"grad_norm": 0.5168038606643677,
|
| 1302 |
+
"learning_rate": 0.0003850421144725844,
|
| 1303 |
+
"loss": 0.6559,
|
| 1304 |
+
"step": 18500
|
| 1305 |
+
},
|
| 1306 |
+
{
|
| 1307 |
+
"epoch": 126.96245733788396,
|
| 1308 |
+
"grad_norm": 0.5486589670181274,
|
| 1309 |
+
"learning_rate": 0.00038039645708488955,
|
| 1310 |
+
"loss": 0.6606,
|
| 1311 |
+
"step": 18600
|
| 1312 |
+
},
|
| 1313 |
+
{
|
| 1314 |
+
"epoch": 127.64505119453925,
|
| 1315 |
+
"grad_norm": 0.44690069556236267,
|
| 1316 |
+
"learning_rate": 0.0003757717742777895,
|
| 1317 |
+
"loss": 0.6538,
|
| 1318 |
+
"step": 18700
|
| 1319 |
+
},
|
| 1320 |
+
{
|
| 1321 |
+
"epoch": 128.32764505119454,
|
| 1322 |
+
"grad_norm": 0.5112203359603882,
|
| 1323 |
+
"learning_rate": 0.00037116863797791145,
|
| 1324 |
+
"loss": 0.6523,
|
| 1325 |
+
"step": 18800
|
| 1326 |
+
},
|
| 1327 |
+
{
|
| 1328 |
+
"epoch": 129.01023890784984,
|
| 1329 |
+
"grad_norm": 0.479837030172348,
|
| 1330 |
+
"learning_rate": 0.0003665876174472628,
|
| 1331 |
+
"loss": 0.6563,
|
| 1332 |
+
"step": 18900
|
| 1333 |
+
},
|
| 1334 |
+
{
|
| 1335 |
+
"epoch": 129.6928327645051,
|
| 1336 |
+
"grad_norm": 0.4698416292667389,
|
| 1337 |
+
"learning_rate": 0.00036202927921283144,
|
| 1338 |
+
"loss": 0.6495,
|
| 1339 |
+
"step": 19000
|
| 1340 |
+
},
|
| 1341 |
+
{
|
| 1342 |
+
"epoch": 130.37542662116041,
|
| 1343 |
+
"grad_norm": 0.5283064246177673,
|
| 1344 |
+
"learning_rate": 0.0003574941869965248,
|
| 1345 |
+
"loss": 0.6474,
|
| 1346 |
+
"step": 19100
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 131.0580204778157,
|
| 1350 |
+
"grad_norm": 0.4855287969112396,
|
| 1351 |
+
"learning_rate": 0.00035298290164545443,
|
| 1352 |
+
"loss": 0.652,
|
| 1353 |
+
"step": 19200
|
| 1354 |
+
},
|
| 1355 |
+
{
|
| 1356 |
+
"epoch": 131.740614334471,
|
| 1357 |
+
"grad_norm": 0.4482058882713318,
|
| 1358 |
+
"learning_rate": 0.00034849598106257795,
|
| 1359 |
+
"loss": 0.6454,
|
| 1360 |
+
"step": 19300
|
| 1361 |
+
},
|
| 1362 |
+
{
|
| 1363 |
+
"epoch": 132.4232081911263,
|
| 1364 |
+
"grad_norm": 0.487212210893631,
|
| 1365 |
+
"learning_rate": 0.00034403398013770347,
|
| 1366 |
+
"loss": 0.643,
|
| 1367 |
+
"step": 19400
|
| 1368 |
+
},
|
| 1369 |
+
{
|
| 1370 |
+
"epoch": 133.10580204778157,
|
| 1371 |
+
"grad_norm": 0.4655930697917938,
|
| 1372 |
+
"learning_rate": 0.0003395974506788674,
|
| 1373 |
+
"loss": 0.6477,
|
| 1374 |
+
"step": 19500
|
| 1375 |
+
},
|
| 1376 |
+
{
|
| 1377 |
+
"epoch": 133.78839590443687,
|
| 1378 |
+
"grad_norm": 0.4315454661846161,
|
| 1379 |
+
"learning_rate": 0.00033518694134409367,
|
| 1380 |
+
"loss": 0.6411,
|
| 1381 |
+
"step": 19600
|
| 1382 |
+
},
|
| 1383 |
+
{
|
| 1384 |
+
"epoch": 134.47098976109214,
|
| 1385 |
+
"grad_norm": 0.505107581615448,
|
| 1386 |
+
"learning_rate": 0.00033080299757354216,
|
| 1387 |
+
"loss": 0.6396,
|
| 1388 |
+
"step": 19700
|
| 1389 |
+
},
|
| 1390 |
+
{
|
| 1391 |
+
"epoch": 135.15358361774744,
|
| 1392 |
+
"grad_norm": 0.47746554017066956,
|
| 1393 |
+
"learning_rate": 0.00032644616152205385,
|
| 1394 |
+
"loss": 0.6436,
|
| 1395 |
+
"step": 19800
|
| 1396 |
+
},
|
| 1397 |
+
{
|
| 1398 |
+
"epoch": 135.83617747440272,
|
| 1399 |
+
"grad_norm": 0.44861286878585815,
|
| 1400 |
+
"learning_rate": 0.0003221169719921052,
|
| 1401 |
+
"loss": 0.6382,
|
| 1402 |
+
"step": 19900
|
| 1403 |
+
},
|
| 1404 |
+
{
|
| 1405 |
+
"epoch": 136.51877133105802,
|
| 1406 |
+
"grad_norm": 0.4924197196960449,
|
| 1407 |
+
"learning_rate": 0.00031781596436717465,
|
| 1408 |
+
"loss": 0.636,
|
| 1409 |
+
"step": 20000
|
| 1410 |
+
},
|
| 1411 |
+
{
|
| 1412 |
+
"epoch": 137.20136518771332,
|
| 1413 |
+
"grad_norm": 0.4533005952835083,
|
| 1414 |
+
"learning_rate": 0.0003135436705455319,
|
| 1415 |
+
"loss": 0.6388,
|
| 1416 |
+
"step": 20100
|
| 1417 |
+
},
|
| 1418 |
+
{
|
| 1419 |
+
"epoch": 137.8839590443686,
|
| 1420 |
+
"grad_norm": 0.4372631311416626,
|
| 1421 |
+
"learning_rate": 0.0003093006188744598,
|
| 1422 |
+
"loss": 0.6349,
|
| 1423 |
+
"step": 20200
|
| 1424 |
+
},
|
| 1425 |
+
{
|
| 1426 |
+
"epoch": 138.5665529010239,
|
| 1427 |
+
"grad_norm": 0.5831769108772278,
|
| 1428 |
+
"learning_rate": 0.0003050873340849148,
|
| 1429 |
+
"loss": 0.6323,
|
| 1430 |
+
"step": 20300
|
| 1431 |
+
},
|
| 1432 |
+
{
|
| 1433 |
+
"epoch": 139.24914675767917,
|
| 1434 |
+
"grad_norm": 0.41959720849990845,
|
| 1435 |
+
"learning_rate": 0.00030090433722663367,
|
| 1436 |
+
"loss": 0.6357,
|
| 1437 |
+
"step": 20400
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 139.93174061433447,
|
| 1441 |
+
"grad_norm": 0.47732993960380554,
|
| 1442 |
+
"learning_rate": 0.00029675214560369637,
|
| 1443 |
+
"loss": 0.6323,
|
| 1444 |
+
"step": 20500
|
| 1445 |
+
},
|
| 1446 |
+
{
|
| 1447 |
+
"epoch": 140.61433447098977,
|
| 1448 |
+
"grad_norm": 0.4794413447380066,
|
| 1449 |
+
"learning_rate": 0.0002926312727105518,
|
| 1450 |
+
"loss": 0.6303,
|
| 1451 |
+
"step": 20600
|
| 1452 |
+
},
|
| 1453 |
+
{
|
| 1454 |
+
"epoch": 141.29692832764505,
|
| 1455 |
+
"grad_norm": 0.42916139960289,
|
| 1456 |
+
"learning_rate": 0.00028854222816851547,
|
| 1457 |
+
"loss": 0.6327,
|
| 1458 |
+
"step": 20700
|
| 1459 |
+
},
|
| 1460 |
+
{
|
| 1461 |
+
"epoch": 141.97952218430035,
|
| 1462 |
+
"grad_norm": 0.42981767654418945,
|
| 1463 |
+
"learning_rate": 0.00028448551766274466,
|
| 1464 |
+
"loss": 0.6285,
|
| 1465 |
+
"step": 20800
|
| 1466 |
+
},
|
| 1467 |
+
{
|
| 1468 |
+
"epoch": 142.66211604095562,
|
| 1469 |
+
"grad_norm": 0.48353445529937744,
|
| 1470 |
+
"learning_rate": 0.0002804616428797021,
|
| 1471 |
+
"loss": 0.6272,
|
| 1472 |
+
"step": 20900
|
| 1473 |
+
},
|
| 1474 |
+
{
|
| 1475 |
+
"epoch": 143.34470989761093,
|
| 1476 |
+
"grad_norm": 0.4736007750034332,
|
| 1477 |
+
"learning_rate": 0.0002764711014451121,
|
| 1478 |
+
"loss": 0.6288,
|
| 1479 |
+
"step": 21000
|
| 1480 |
+
},
|
| 1481 |
+
{
|
| 1482 |
+
"epoch": 144.0273037542662,
|
| 1483 |
+
"grad_norm": 0.47045692801475525,
|
| 1484 |
+
"learning_rate": 0.0002725143868624209,
|
| 1485 |
+
"loss": 0.6257,
|
| 1486 |
+
"step": 21100
|
| 1487 |
+
},
|
| 1488 |
+
{
|
| 1489 |
+
"epoch": 144.7098976109215,
|
| 1490 |
+
"grad_norm": 0.4963953495025635,
|
| 1491 |
+
"learning_rate": 0.0002685919884517659,
|
| 1492 |
+
"loss": 0.6251,
|
| 1493 |
+
"step": 21200
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"epoch": 145.3924914675768,
|
| 1497 |
+
"grad_norm": 0.4388594925403595,
|
| 1498 |
+
"learning_rate": 0.00026470439128946214,
|
| 1499 |
+
"loss": 0.6249,
|
| 1500 |
+
"step": 21300
|
| 1501 |
+
},
|
| 1502 |
+
{
|
| 1503 |
+
"epoch": 146.07508532423208,
|
| 1504 |
+
"grad_norm": 0.4757547378540039,
|
| 1505 |
+
"learning_rate": 0.0002608520761480135,
|
| 1506 |
+
"loss": 0.6241,
|
| 1507 |
+
"step": 21400
|
| 1508 |
+
},
|
| 1509 |
+
{
|
| 1510 |
+
"epoch": 146.75767918088738,
|
| 1511 |
+
"grad_norm": 0.46713003516197205,
|
| 1512 |
+
"learning_rate": 0.00025703551943665636,
|
| 1513 |
+
"loss": 0.6229,
|
| 1514 |
+
"step": 21500
|
| 1515 |
+
},
|
| 1516 |
+
{
|
| 1517 |
+
"epoch": 147.44027303754265,
|
| 1518 |
+
"grad_norm": 0.43607574701309204,
|
| 1519 |
+
"learning_rate": 0.00025325519314244374,
|
| 1520 |
+
"loss": 0.623,
|
| 1521 |
+
"step": 21600
|
| 1522 |
+
},
|
| 1523 |
+
{
|
| 1524 |
+
"epoch": 148.12286689419795,
|
| 1525 |
+
"grad_norm": 0.4766788184642792,
|
| 1526 |
+
"learning_rate": 0.00024951156477187413,
|
| 1527 |
+
"loss": 0.6212,
|
| 1528 |
+
"step": 21700
|
| 1529 |
+
},
|
| 1530 |
+
{
|
| 1531 |
+
"epoch": 148.80546075085323,
|
| 1532 |
+
"grad_norm": 0.49822133779525757,
|
| 1533 |
+
"learning_rate": 0.00024580509729307673,
|
| 1534 |
+
"loss": 0.6215,
|
| 1535 |
+
"step": 21800
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"epoch": 149.48805460750853,
|
| 1539 |
+
"grad_norm": 0.4069366455078125,
|
| 1540 |
+
"learning_rate": 0.0002421362490785569,
|
| 1541 |
+
"loss": 0.6199,
|
| 1542 |
+
"step": 21900
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 150.17064846416383,
|
| 1546 |
+
"grad_norm": 0.46552979946136475,
|
| 1547 |
+
"learning_rate": 0.00023850547384850974,
|
| 1548 |
+
"loss": 0.6194,
|
| 1549 |
+
"step": 22000
|
| 1550 |
+
},
|
| 1551 |
+
{
|
| 1552 |
+
"epoch": 150.8532423208191,
|
| 1553 |
+
"grad_norm": 0.45079755783081055,
|
| 1554 |
+
"learning_rate": 0.00023491322061470915,
|
| 1555 |
+
"loss": 0.6202,
|
| 1556 |
+
"step": 22100
|
| 1557 |
+
},
|
| 1558 |
+
{
|
| 1559 |
+
"epoch": 151.5358361774744,
|
| 1560 |
+
"grad_norm": 0.46759602427482605,
|
| 1561 |
+
"learning_rate": 0.00023135993362497948,
|
| 1562 |
+
"loss": 0.6181,
|
| 1563 |
+
"step": 22200
|
| 1564 |
+
},
|
| 1565 |
+
{
|
| 1566 |
+
"epoch": 152.21843003412968,
|
| 1567 |
+
"grad_norm": 0.45771974325180054,
|
| 1568 |
+
"learning_rate": 0.00022784605230825666,
|
| 1569 |
+
"loss": 0.6164,
|
| 1570 |
+
"step": 22300
|
| 1571 |
+
},
|
| 1572 |
+
{
|
| 1573 |
+
"epoch": 152.90102389078498,
|
| 1574 |
+
"grad_norm": 0.5069631338119507,
|
| 1575 |
+
"learning_rate": 0.00022437201122024375,
|
| 1576 |
+
"loss": 0.6187,
|
| 1577 |
+
"step": 22400
|
| 1578 |
+
},
|
| 1579 |
+
{
|
| 1580 |
+
"epoch": 153.5836177474403,
|
| 1581 |
+
"grad_norm": 0.43779122829437256,
|
| 1582 |
+
"learning_rate": 0.00022093823998967115,
|
| 1583 |
+
"loss": 0.6164,
|
| 1584 |
+
"step": 22500
|
| 1585 |
+
},
|
| 1586 |
+
{
|
| 1587 |
+
"epoch": 154.26621160409556,
|
| 1588 |
+
"grad_norm": 0.477224200963974,
|
| 1589 |
+
"learning_rate": 0.00021754516326516458,
|
| 1590 |
+
"loss": 0.6151,
|
| 1591 |
+
"step": 22600
|
| 1592 |
+
},
|
| 1593 |
+
{
|
| 1594 |
+
"epoch": 154.94880546075086,
|
| 1595 |
+
"grad_norm": 0.47925087809562683,
|
| 1596 |
+
"learning_rate": 0.00021419320066272973,
|
| 1597 |
+
"loss": 0.6166,
|
| 1598 |
+
"step": 22700
|
| 1599 |
+
},
|
| 1600 |
+
{
|
| 1601 |
+
"epoch": 155.63139931740614,
|
| 1602 |
+
"grad_norm": 0.4455689787864685,
|
| 1603 |
+
"learning_rate": 0.00021088276671385882,
|
| 1604 |
+
"loss": 0.6139,
|
| 1605 |
+
"step": 22800
|
| 1606 |
+
},
|
| 1607 |
+
{
|
| 1608 |
+
"epoch": 156.31399317406144,
|
| 1609 |
+
"grad_norm": 0.4499022364616394,
|
| 1610 |
+
"learning_rate": 0.00020761427081426677,
|
| 1611 |
+
"loss": 0.6132,
|
| 1612 |
+
"step": 22900
|
| 1613 |
+
},
|
| 1614 |
+
{
|
| 1615 |
+
"epoch": 156.9965870307167,
|
| 1616 |
+
"grad_norm": 0.4519612193107605,
|
| 1617 |
+
"learning_rate": 0.00020438811717326112,
|
| 1618 |
+
"loss": 0.6154,
|
| 1619 |
+
"step": 23000
|
| 1620 |
+
},
|
| 1621 |
+
{
|
| 1622 |
+
"epoch": 157.679180887372,
|
| 1623 |
+
"grad_norm": 0.4130764603614807,
|
| 1624 |
+
"learning_rate": 0.0002012047047637547,
|
| 1625 |
+
"loss": 0.6122,
|
| 1626 |
+
"step": 23100
|
| 1627 |
+
},
|
| 1628 |
+
{
|
| 1629 |
+
"epoch": 158.36177474402731,
|
| 1630 |
+
"grad_norm": 0.44488415122032166,
|
| 1631 |
+
"learning_rate": 0.0001980644272729254,
|
| 1632 |
+
"loss": 0.6132,
|
| 1633 |
+
"step": 23200
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 159.0443686006826,
|
| 1637 |
+
"grad_norm": 0.47042757272720337,
|
| 1638 |
+
"learning_rate": 0.00019496767305352946,
|
| 1639 |
+
"loss": 0.6135,
|
| 1640 |
+
"step": 23300
|
| 1641 |
+
},
|
| 1642 |
+
{
|
| 1643 |
+
"epoch": 159.7269624573379,
|
| 1644 |
+
"grad_norm": 0.3974725604057312,
|
| 1645 |
+
"learning_rate": 0.00019191482507587376,
|
| 1646 |
+
"loss": 0.6107,
|
| 1647 |
+
"step": 23400
|
| 1648 |
+
},
|
| 1649 |
+
{
|
| 1650 |
+
"epoch": 160.40955631399316,
|
| 1651 |
+
"grad_norm": 0.46765318512916565,
|
| 1652 |
+
"learning_rate": 0.00018890626088045593,
|
| 1653 |
+
"loss": 0.6109,
|
| 1654 |
+
"step": 23500
|
| 1655 |
+
},
|
| 1656 |
+
{
|
| 1657 |
+
"epoch": 161.09215017064847,
|
| 1658 |
+
"grad_norm": 0.42309921979904175,
|
| 1659 |
+
"learning_rate": 0.00018594235253127373,
|
| 1660 |
+
"loss": 0.6122,
|
| 1661 |
+
"step": 23600
|
| 1662 |
+
},
|
| 1663 |
+
{
|
| 1664 |
+
"epoch": 161.77474402730377,
|
| 1665 |
+
"grad_norm": 0.3989265561103821,
|
| 1666 |
+
"learning_rate": 0.00018302346656981234,
|
| 1667 |
+
"loss": 0.6099,
|
| 1668 |
+
"step": 23700
|
| 1669 |
+
},
|
| 1670 |
+
{
|
| 1671 |
+
"epoch": 162.45733788395904,
|
| 1672 |
+
"grad_norm": 0.4663792550563812,
|
| 1673 |
+
"learning_rate": 0.0001801499639697153,
|
| 1674 |
+
"loss": 0.6103,
|
| 1675 |
+
"step": 23800
|
| 1676 |
+
},
|
| 1677 |
+
{
|
| 1678 |
+
"epoch": 163.13993174061434,
|
| 1679 |
+
"grad_norm": 0.3913436233997345,
|
| 1680 |
+
"learning_rate": 0.0001773222000921435,
|
| 1681 |
+
"loss": 0.6106,
|
| 1682 |
+
"step": 23900
|
| 1683 |
+
},
|
| 1684 |
+
{
|
| 1685 |
+
"epoch": 163.82252559726962,
|
| 1686 |
+
"grad_norm": 0.41263365745544434,
|
| 1687 |
+
"learning_rate": 0.00017454052464182778,
|
| 1688 |
+
"loss": 0.6092,
|
| 1689 |
+
"step": 24000
|
| 1690 |
+
},
|
| 1691 |
+
{
|
| 1692 |
+
"epoch": 164.50511945392492,
|
| 1693 |
+
"grad_norm": 0.4633448123931885,
|
| 1694 |
+
"learning_rate": 0.0001718052816238222,
|
| 1695 |
+
"loss": 0.608,
|
| 1696 |
+
"step": 24100
|
| 1697 |
+
},
|
| 1698 |
+
{
|
| 1699 |
+
"epoch": 165.1877133105802,
|
| 1700 |
+
"grad_norm": 0.3786383271217346,
|
| 1701 |
+
"learning_rate": 0.00016911680930096085,
|
| 1702 |
+
"loss": 0.6094,
|
| 1703 |
+
"step": 24200
|
| 1704 |
+
},
|
| 1705 |
+
{
|
| 1706 |
+
"epoch": 165.8703071672355,
|
| 1707 |
+
"grad_norm": 0.4065646529197693,
|
| 1708 |
+
"learning_rate": 0.00016647544015202623,
|
| 1709 |
+
"loss": 0.608,
|
| 1710 |
+
"step": 24300
|
| 1711 |
+
},
|
| 1712 |
+
{
|
| 1713 |
+
"epoch": 166.5529010238908,
|
| 1714 |
+
"grad_norm": 0.43729546666145325,
|
| 1715 |
+
"learning_rate": 0.00016388150083063117,
|
| 1716 |
+
"loss": 0.6072,
|
| 1717 |
+
"step": 24400
|
| 1718 |
+
},
|
| 1719 |
+
{
|
| 1720 |
+
"epoch": 167.23549488054607,
|
| 1721 |
+
"grad_norm": 0.4111308157444,
|
| 1722 |
+
"learning_rate": 0.00016133531212482325,
|
| 1723 |
+
"loss": 0.6078,
|
| 1724 |
+
"step": 24500
|
| 1725 |
+
},
|
| 1726 |
+
{
|
| 1727 |
+
"epoch": 167.91808873720137,
|
| 1728 |
+
"grad_norm": 0.39553767442703247,
|
| 1729 |
+
"learning_rate": 0.00015883718891741258,
|
| 1730 |
+
"loss": 0.6066,
|
| 1731 |
+
"step": 24600
|
| 1732 |
+
},
|
| 1733 |
+
{
|
| 1734 |
+
"epoch": 168.60068259385665,
|
| 1735 |
+
"grad_norm": 0.4074500501155853,
|
| 1736 |
+
"learning_rate": 0.00015638744014703128,
|
| 1737 |
+
"loss": 0.6068,
|
| 1738 |
+
"step": 24700
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 169.28327645051195,
|
| 1742 |
+
"grad_norm": 0.42327943444252014,
|
| 1743 |
+
"learning_rate": 0.0001539863687699274,
|
| 1744 |
+
"loss": 0.607,
|
| 1745 |
+
"step": 24800
|
| 1746 |
+
},
|
| 1747 |
+
{
|
| 1748 |
+
"epoch": 169.96587030716722,
|
| 1749 |
+
"grad_norm": 0.40863099694252014,
|
| 1750 |
+
"learning_rate": 0.0001516342717224992,
|
| 1751 |
+
"loss": 0.6067,
|
| 1752 |
+
"step": 24900
|
| 1753 |
+
},
|
| 1754 |
+
{
|
| 1755 |
+
"epoch": 170.64846416382252,
|
| 1756 |
+
"grad_norm": 0.4853709638118744,
|
| 1757 |
+
"learning_rate": 0.00014933143988457237,
|
| 1758 |
+
"loss": 0.6056,
|
| 1759 |
+
"step": 25000
|
| 1760 |
+
},
|
| 1761 |
+
{
|
| 1762 |
+
"epoch": 171.33105802047783,
|
| 1763 |
+
"grad_norm": 0.383801132440567,
|
| 1764 |
+
"learning_rate": 0.00014707815804342907,
|
| 1765 |
+
"loss": 0.6063,
|
| 1766 |
+
"step": 25100
|
| 1767 |
+
},
|
| 1768 |
+
{
|
| 1769 |
+
"epoch": 172.0136518771331,
|
| 1770 |
+
"grad_norm": 0.4241873621940613,
|
| 1771 |
+
"learning_rate": 0.00014487470485858767,
|
| 1772 |
+
"loss": 0.6058,
|
| 1773 |
+
"step": 25200
|
| 1774 |
+
},
|
| 1775 |
+
{
|
| 1776 |
+
"epoch": 172.6962457337884,
|
| 1777 |
+
"grad_norm": 0.4262351989746094,
|
| 1778 |
+
"learning_rate": 0.00014272135282734167,
|
| 1779 |
+
"loss": 0.6043,
|
| 1780 |
+
"step": 25300
|
| 1781 |
+
},
|
| 1782 |
+
{
|
| 1783 |
+
"epoch": 173.37883959044368,
|
| 1784 |
+
"grad_norm": 0.38850197196006775,
|
| 1785 |
+
"learning_rate": 0.00014061836825106022,
|
| 1786 |
+
"loss": 0.606,
|
| 1787 |
+
"step": 25400
|
| 1788 |
+
},
|
| 1789 |
+
{
|
| 1790 |
+
"epoch": 174.06143344709898,
|
| 1791 |
+
"grad_norm": 0.4169691801071167,
|
| 1792 |
+
"learning_rate": 0.00013856601120225574,
|
| 1793 |
+
"loss": 0.6038,
|
| 1794 |
+
"step": 25500
|
| 1795 |
+
},
|
| 1796 |
+
{
|
| 1797 |
+
"epoch": 174.74402730375428,
|
| 1798 |
+
"grad_norm": 0.43140584230422974,
|
| 1799 |
+
"learning_rate": 0.00013656453549242055,
|
| 1800 |
+
"loss": 0.6043,
|
| 1801 |
+
"step": 25600
|
| 1802 |
+
},
|
| 1803 |
+
{
|
| 1804 |
+
"epoch": 175.42662116040955,
|
| 1805 |
+
"grad_norm": 0.3729684054851532,
|
| 1806 |
+
"learning_rate": 0.00013461418864063861,
|
| 1807 |
+
"loss": 0.6047,
|
| 1808 |
+
"step": 25700
|
| 1809 |
+
},
|
| 1810 |
+
{
|
| 1811 |
+
"epoch": 176.10921501706486,
|
| 1812 |
+
"grad_norm": 0.435611367225647,
|
| 1813 |
+
"learning_rate": 0.00013271521184297523,
|
| 1814 |
+
"loss": 0.6035,
|
| 1815 |
+
"step": 25800
|
| 1816 |
+
},
|
| 1817 |
+
{
|
| 1818 |
+
"epoch": 176.79180887372013,
|
| 1819 |
+
"grad_norm": 0.4359755218029022,
|
| 1820 |
+
"learning_rate": 0.00013086783994264905,
|
| 1821 |
+
"loss": 0.6044,
|
| 1822 |
+
"step": 25900
|
| 1823 |
+
},
|
| 1824 |
+
{
|
| 1825 |
+
"epoch": 177.47440273037543,
|
| 1826 |
+
"grad_norm": 0.3867483139038086,
|
| 1827 |
+
"learning_rate": 0.00012907230140098895,
|
| 1828 |
+
"loss": 0.6031,
|
| 1829 |
+
"step": 26000
|
| 1830 |
+
},
|
| 1831 |
+
{
|
| 1832 |
+
"epoch": 178.1569965870307,
|
| 1833 |
+
"grad_norm": 0.41799867153167725,
|
| 1834 |
+
"learning_rate": 0.00012732881826918094,
|
| 1835 |
+
"loss": 0.6031,
|
| 1836 |
+
"step": 26100
|
| 1837 |
+
},
|
| 1838 |
+
{
|
| 1839 |
+
"epoch": 178.839590443686,
|
| 1840 |
+
"grad_norm": 0.43031203746795654,
|
| 1841 |
+
"learning_rate": 0.00012563760616080745,
|
| 1842 |
+
"loss": 0.6028,
|
| 1843 |
+
"step": 26200
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 179.5221843003413,
|
| 1847 |
+
"grad_norm": 0.3603518605232239,
|
| 1848 |
+
"learning_rate": 0.00012399887422518243,
|
| 1849 |
+
"loss": 0.6026,
|
| 1850 |
+
"step": 26300
|
| 1851 |
+
},
|
| 1852 |
+
{
|
| 1853 |
+
"epoch": 180.20477815699658,
|
| 1854 |
+
"grad_norm": 0.42124807834625244,
|
| 1855 |
+
"learning_rate": 0.0001224128251214867,
|
| 1856 |
+
"loss": 0.6021,
|
| 1857 |
+
"step": 26400
|
| 1858 |
+
},
|
| 1859 |
+
{
|
| 1860 |
+
"epoch": 180.88737201365188,
|
| 1861 |
+
"grad_norm": 0.40946000814437866,
|
| 1862 |
+
"learning_rate": 0.00012087965499370513,
|
| 1863 |
+
"loss": 0.6036,
|
| 1864 |
+
"step": 26500
|
| 1865 |
+
},
|
| 1866 |
+
{
|
| 1867 |
+
"epoch": 181.56996587030716,
|
| 1868 |
+
"grad_norm": 0.3763185143470764,
|
| 1869 |
+
"learning_rate": 0.00011939955344636983,
|
| 1870 |
+
"loss": 0.6022,
|
| 1871 |
+
"step": 26600
|
| 1872 |
+
},
|
| 1873 |
+
{
|
| 1874 |
+
"epoch": 182.25255972696246,
|
| 1875 |
+
"grad_norm": 0.4167167544364929,
|
| 1876 |
+
"learning_rate": 0.00011797270352111224,
|
| 1877 |
+
"loss": 0.6016,
|
| 1878 |
+
"step": 26700
|
| 1879 |
+
},
|
| 1880 |
+
{
|
| 1881 |
+
"epoch": 182.93515358361773,
|
| 1882 |
+
"grad_norm": 0.42429491877555847,
|
| 1883 |
+
"learning_rate": 0.0001165992816740268,
|
| 1884 |
+
"loss": 0.6021,
|
| 1885 |
+
"step": 26800
|
| 1886 |
+
},
|
| 1887 |
+
{
|
| 1888 |
+
"epoch": 183.61774744027304,
|
| 1889 |
+
"grad_norm": 0.38188785314559937,
|
| 1890 |
+
"learning_rate": 0.00011527945775384831,
|
| 1891 |
+
"loss": 0.6022,
|
| 1892 |
+
"step": 26900
|
| 1893 |
+
},
|
| 1894 |
+
{
|
| 1895 |
+
"epoch": 184.30034129692834,
|
| 1896 |
+
"grad_norm": 0.47042107582092285,
|
| 1897 |
+
"learning_rate": 0.00011401339498094774,
|
| 1898 |
+
"loss": 0.6016,
|
| 1899 |
+
"step": 27000
|
| 1900 |
+
},
|
| 1901 |
+
{
|
| 1902 |
+
"epoch": 184.9829351535836,
|
| 1903 |
+
"grad_norm": 0.422690212726593,
|
| 1904 |
+
"learning_rate": 0.00011280124992714677,
|
| 1905 |
+
"loss": 0.602,
|
| 1906 |
+
"step": 27100
|
| 1907 |
+
},
|
| 1908 |
+
{
|
| 1909 |
+
"epoch": 185.6655290102389,
|
| 1910 |
+
"grad_norm": 0.3714023232460022,
|
| 1911 |
+
"learning_rate": 0.00011164317249635475,
|
| 1912 |
+
"loss": 0.6014,
|
| 1913 |
+
"step": 27200
|
| 1914 |
+
},
|
| 1915 |
+
{
|
| 1916 |
+
"epoch": 186.3481228668942,
|
| 1917 |
+
"grad_norm": 0.43933653831481934,
|
| 1918 |
+
"learning_rate": 0.00011053930590603041,
|
| 1919 |
+
"loss": 0.6003,
|
| 1920 |
+
"step": 27300
|
| 1921 |
+
},
|
| 1922 |
+
{
|
| 1923 |
+
"epoch": 187.0307167235495,
|
| 1924 |
+
"grad_norm": 0.3583015501499176,
|
| 1925 |
+
"learning_rate": 0.00010948978666947039,
|
| 1926 |
+
"loss": 0.6016,
|
| 1927 |
+
"step": 27400
|
| 1928 |
+
},
|
| 1929 |
+
{
|
| 1930 |
+
"epoch": 187.7133105802048,
|
| 1931 |
+
"grad_norm": 0.38929620385169983,
|
| 1932 |
+
"learning_rate": 0.00010849474457892682,
|
| 1933 |
+
"loss": 0.6006,
|
| 1934 |
+
"step": 27500
|
| 1935 |
+
},
|
| 1936 |
+
{
|
| 1937 |
+
"epoch": 188.39590443686006,
|
| 1938 |
+
"grad_norm": 0.42569249868392944,
|
| 1939 |
+
"learning_rate": 0.0001075543026895563,
|
| 1940 |
+
"loss": 0.6008,
|
| 1941 |
+
"step": 27600
|
| 1942 |
+
},
|
| 1943 |
+
{
|
| 1944 |
+
"epoch": 189.07849829351537,
|
| 1945 |
+
"grad_norm": 0.37430456280708313,
|
| 1946 |
+
"learning_rate": 0.00010666857730420167,
|
| 1947 |
+
"loss": 0.601,
|
| 1948 |
+
"step": 27700
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 189.76109215017064,
|
| 1952 |
+
"grad_norm": 0.4152965545654297,
|
| 1953 |
+
"learning_rate": 0.00010583767795900921,
|
| 1954 |
+
"loss": 0.6006,
|
| 1955 |
+
"step": 27800
|
| 1956 |
+
},
|
| 1957 |
+
{
|
| 1958 |
+
"epoch": 190.44368600682594,
|
| 1959 |
+
"grad_norm": 0.4082518517971039,
|
| 1960 |
+
"learning_rate": 0.00010506170740988232,
|
| 1961 |
+
"loss": 0.5998,
|
| 1962 |
+
"step": 27900
|
| 1963 |
+
},
|
| 1964 |
+
{
|
| 1965 |
+
"epoch": 191.12627986348122,
|
| 1966 |
+
"grad_norm": 0.3863820731639862,
|
| 1967 |
+
"learning_rate": 0.00010434076161977406,
|
| 1968 |
+
"loss": 0.6013,
|
| 1969 |
+
"step": 28000
|
| 1970 |
+
},
|
| 1971 |
+
{
|
| 1972 |
+
"epoch": 191.80887372013652,
|
| 1973 |
+
"grad_norm": 0.36929044127464294,
|
| 1974 |
+
"learning_rate": 0.00010367492974681949,
|
| 1975 |
+
"loss": 0.6007,
|
| 1976 |
+
"step": 28100
|
| 1977 |
+
},
|
| 1978 |
+
{
|
| 1979 |
+
"epoch": 192.49146757679182,
|
| 1980 |
+
"grad_norm": 0.3993302583694458,
|
| 1981 |
+
"learning_rate": 0.0001030642941333095,
|
| 1982 |
+
"loss": 0.6,
|
| 1983 |
+
"step": 28200
|
| 1984 |
+
},
|
| 1985 |
+
{
|
| 1986 |
+
"epoch": 193.1740614334471,
|
| 1987 |
+
"grad_norm": 0.36054790019989014,
|
| 1988 |
+
"learning_rate": 0.00010250893029550792,
|
| 1989 |
+
"loss": 0.6006,
|
| 1990 |
+
"step": 28300
|
| 1991 |
+
},
|
| 1992 |
+
{
|
| 1993 |
+
"epoch": 193.8566552901024,
|
| 1994 |
+
"grad_norm": 0.40515759587287903,
|
| 1995 |
+
"learning_rate": 0.00010200890691431253,
|
| 1996 |
+
"loss": 0.6004,
|
| 1997 |
+
"step": 28400
|
| 1998 |
+
},
|
| 1999 |
+
{
|
| 2000 |
+
"epoch": 194.53924914675767,
|
| 2001 |
+
"grad_norm": 0.46927398443222046,
|
| 2002 |
+
"learning_rate": 0.00010156428582676107,
|
| 2003 |
+
"loss": 0.5991,
|
| 2004 |
+
"step": 28500
|
| 2005 |
+
},
|
| 2006 |
+
{
|
| 2007 |
+
"epoch": 195.22184300341297,
|
| 2008 |
+
"grad_norm": 0.3784146308898926,
|
| 2009 |
+
"learning_rate": 0.00010117512201838442,
|
| 2010 |
+
"loss": 0.6005,
|
| 2011 |
+
"step": 28600
|
| 2012 |
+
},
|
| 2013 |
+
{
|
| 2014 |
+
"epoch": 195.90443686006827,
|
| 2015 |
+
"grad_norm": 0.3883312940597534,
|
| 2016 |
+
"learning_rate": 0.00010084146361640634,
|
| 2017 |
+
"loss": 0.5991,
|
| 2018 |
+
"step": 28700
|
| 2019 |
+
},
|
| 2020 |
+
{
|
| 2021 |
+
"epoch": 196.58703071672355,
|
| 2022 |
+
"grad_norm": 0.42353108525276184,
|
| 2023 |
+
"learning_rate": 0.0001005633518837918,
|
| 2024 |
+
"loss": 0.5996,
|
| 2025 |
+
"step": 28800
|
| 2026 |
+
},
|
| 2027 |
+
{
|
| 2028 |
+
"epoch": 197.26962457337885,
|
| 2029 |
+
"grad_norm": 0.4231187701225281,
|
| 2030 |
+
"learning_rate": 0.00010034082121414392,
|
| 2031 |
+
"loss": 0.5996,
|
| 2032 |
+
"step": 28900
|
| 2033 |
+
},
|
| 2034 |
+
{
|
| 2035 |
+
"epoch": 197.95221843003412,
|
| 2036 |
+
"grad_norm": 0.3672020733356476,
|
| 2037 |
+
"learning_rate": 0.0001001738991274507,
|
| 2038 |
+
"loss": 0.6002,
|
| 2039 |
+
"step": 29000
|
| 2040 |
+
}
|
| 2041 |
+
],
|
| 2042 |
+
"logging_steps": 100,
|
| 2043 |
+
"max_steps": 29200,
|
| 2044 |
+
"num_input_tokens_seen": 0,
|
| 2045 |
+
"num_train_epochs": 200,
|
| 2046 |
+
"save_steps": 500,
|
| 2047 |
+
"stateful_callbacks": {
|
| 2048 |
+
"TrainerControl": {
|
| 2049 |
+
"args": {
|
| 2050 |
+
"should_epoch_stop": false,
|
| 2051 |
+
"should_evaluate": false,
|
| 2052 |
+
"should_log": false,
|
| 2053 |
+
"should_save": true,
|
| 2054 |
+
"should_training_stop": false
|
| 2055 |
+
},
|
| 2056 |
+
"attributes": {}
|
| 2057 |
+
}
|
| 2058 |
+
},
|
| 2059 |
+
"total_flos": 9.277077145979453e+17,
|
| 2060 |
+
"train_batch_size": 256,
|
| 2061 |
+
"trial_name": null,
|
| 2062 |
+
"trial_params": null
|
| 2063 |
+
}
|