Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- README.md +60 -0
- added_tokens.json +28 -0
- all_results.json +8 -0
- config.json +38 -0
- generation_config.json +6 -0
- merges.txt +0 -0
- model-00001-of-00013.safetensors +3 -0
- model-00002-of-00013.safetensors +3 -0
- model-00003-of-00013.safetensors +3 -0
- model-00004-of-00013.safetensors +3 -0
- model-00005-of-00013.safetensors +3 -0
- model-00006-of-00013.safetensors +3 -0
- model-00007-of-00013.safetensors +3 -0
- model-00008-of-00013.safetensors +3 -0
- model-00009-of-00013.safetensors +3 -0
- model-00010-of-00013.safetensors +3 -0
- model-00011-of-00013.safetensors +3 -0
- model-00012-of-00013.safetensors +3 -0
- model-00013-of-00013.safetensors +3 -0
- model.safetensors.index.json +0 -0
- special_tokens_map.json +31 -0
- tokenizer.json +3 -0
- tokenizer_config.json +241 -0
- train_results.json +8 -0
- trainer_log.jsonl +316 -0
- trainer_state.json +2248 -0
- training_args.bin +3 -0
- training_loss.png +0 -0
- vocab.json +0 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
README.md
ADDED
|
@@ -0,0 +1,60 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
library_name: transformers
|
| 3 |
+
license: other
|
| 4 |
+
base_model: Qwen/Qwen3-30B-A3B-Base
|
| 5 |
+
tags:
|
| 6 |
+
- llama-factory
|
| 7 |
+
- full
|
| 8 |
+
- generated_from_trainer
|
| 9 |
+
model-index:
|
| 10 |
+
- name: base_s1k_bs16_lr1e-5_zero3
|
| 11 |
+
results: []
|
| 12 |
+
---
|
| 13 |
+
|
| 14 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
| 15 |
+
should probably proofread and complete it, then remove this comment. -->
|
| 16 |
+
|
| 17 |
+
# base_s1k_bs16_lr1e-5_zero3
|
| 18 |
+
|
| 19 |
+
This model is a fine-tuned version of [Qwen/Qwen3-30B-A3B-Base](https://huggingface.co/Qwen/Qwen3-30B-A3B-Base) on the s1k_1.1 dataset.
|
| 20 |
+
|
| 21 |
+
## Model description
|
| 22 |
+
|
| 23 |
+
More information needed
|
| 24 |
+
|
| 25 |
+
## Intended uses & limitations
|
| 26 |
+
|
| 27 |
+
More information needed
|
| 28 |
+
|
| 29 |
+
## Training and evaluation data
|
| 30 |
+
|
| 31 |
+
More information needed
|
| 32 |
+
|
| 33 |
+
## Training procedure
|
| 34 |
+
|
| 35 |
+
### Training hyperparameters
|
| 36 |
+
|
| 37 |
+
The following hyperparameters were used during training:
|
| 38 |
+
- learning_rate: 1e-05
|
| 39 |
+
- train_batch_size: 2
|
| 40 |
+
- eval_batch_size: 8
|
| 41 |
+
- seed: 42
|
| 42 |
+
- distributed_type: multi-GPU
|
| 43 |
+
- num_devices: 8
|
| 44 |
+
- total_train_batch_size: 16
|
| 45 |
+
- total_eval_batch_size: 64
|
| 46 |
+
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
|
| 47 |
+
- lr_scheduler_type: cosine
|
| 48 |
+
- lr_scheduler_warmup_ratio: 0.1
|
| 49 |
+
- num_epochs: 5.0
|
| 50 |
+
|
| 51 |
+
### Training results
|
| 52 |
+
|
| 53 |
+
|
| 54 |
+
|
| 55 |
+
### Framework versions
|
| 56 |
+
|
| 57 |
+
- Transformers 4.51.3
|
| 58 |
+
- Pytorch 2.6.0+cu124
|
| 59 |
+
- Datasets 3.5.0
|
| 60 |
+
- Tokenizers 0.21.1
|
added_tokens.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</think>": 151668,
|
| 3 |
+
"</tool_call>": 151658,
|
| 4 |
+
"</tool_response>": 151666,
|
| 5 |
+
"<think>": 151667,
|
| 6 |
+
"<tool_call>": 151657,
|
| 7 |
+
"<tool_response>": 151665,
|
| 8 |
+
"<|box_end|>": 151649,
|
| 9 |
+
"<|box_start|>": 151648,
|
| 10 |
+
"<|endoftext|>": 151643,
|
| 11 |
+
"<|file_sep|>": 151664,
|
| 12 |
+
"<|fim_middle|>": 151660,
|
| 13 |
+
"<|fim_pad|>": 151662,
|
| 14 |
+
"<|fim_prefix|>": 151659,
|
| 15 |
+
"<|fim_suffix|>": 151661,
|
| 16 |
+
"<|im_end|>": 151645,
|
| 17 |
+
"<|im_start|>": 151644,
|
| 18 |
+
"<|image_pad|>": 151655,
|
| 19 |
+
"<|object_ref_end|>": 151647,
|
| 20 |
+
"<|object_ref_start|>": 151646,
|
| 21 |
+
"<|quad_end|>": 151651,
|
| 22 |
+
"<|quad_start|>": 151650,
|
| 23 |
+
"<|repo_name|>": 151663,
|
| 24 |
+
"<|video_pad|>": 151656,
|
| 25 |
+
"<|vision_end|>": 151653,
|
| 26 |
+
"<|vision_pad|>": 151654,
|
| 27 |
+
"<|vision_start|>": 151652
|
| 28 |
+
}
|
all_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 78480301031424.0,
|
| 4 |
+
"train_loss": 0.7362074042123462,
|
| 5 |
+
"train_runtime": 9541.1641,
|
| 6 |
+
"train_samples_per_second": 0.524,
|
| 7 |
+
"train_steps_per_second": 0.033
|
| 8 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,38 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3MoeForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"decoder_sparse_step": 1,
|
| 9 |
+
"eos_token_id": 151643,
|
| 10 |
+
"head_dim": 128,
|
| 11 |
+
"hidden_act": "silu",
|
| 12 |
+
"hidden_size": 2048,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 6144,
|
| 15 |
+
"max_position_embeddings": 32768,
|
| 16 |
+
"max_window_layers": 48,
|
| 17 |
+
"mlp_only_layers": [],
|
| 18 |
+
"model_type": "qwen3_moe",
|
| 19 |
+
"moe_intermediate_size": 768,
|
| 20 |
+
"norm_topk_prob": true,
|
| 21 |
+
"num_attention_heads": 32,
|
| 22 |
+
"num_experts": 128,
|
| 23 |
+
"num_experts_per_tok": 8,
|
| 24 |
+
"num_hidden_layers": 48,
|
| 25 |
+
"num_key_value_heads": 4,
|
| 26 |
+
"output_router_logits": false,
|
| 27 |
+
"rms_norm_eps": 1e-06,
|
| 28 |
+
"rope_scaling": null,
|
| 29 |
+
"rope_theta": 1000000.0,
|
| 30 |
+
"router_aux_loss_coef": 0.001,
|
| 31 |
+
"sliding_window": null,
|
| 32 |
+
"tie_word_embeddings": false,
|
| 33 |
+
"torch_dtype": "bfloat16",
|
| 34 |
+
"transformers_version": "4.51.3",
|
| 35 |
+
"use_cache": false,
|
| 36 |
+
"use_sliding_window": false,
|
| 37 |
+
"vocab_size": 151936
|
| 38 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"eos_token_id": 151643,
|
| 4 |
+
"max_new_tokens": 2048,
|
| 5 |
+
"transformers_version": "4.51.3"
|
| 6 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model-00001-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9cb2bf954f69092230073973b1f6179e8d7bcbbeb24e7b5059c07f69d96fe866
|
| 3 |
+
size 4997184968
|
model-00002-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6b1b802683e1a6b83b28c5b7ecc71e10fbe0af0ef8d7f19f3f88681e53f56978
|
| 3 |
+
size 4997741608
|
model-00003-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d6e932846c77535ec74d02d17860fb8e62cbb78625081e14f18e7e5a9b3149c4
|
| 3 |
+
size 4997742208
|
model-00004-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e40cf0b7b93391a9d70f6a199b4f26ecaf18790e472b1f69d5160d340e9e1f7f
|
| 3 |
+
size 4997743184
|
model-00005-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:074e5808637fe389cc1ad0568dd91bba2b11590178b8ca435295128b6c5238b2
|
| 3 |
+
size 4997743184
|
model-00006-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:771fe3cf84f7d9aebd6d2f9413f26c9aacf1a81cfca6996ec1acc6537d21c909
|
| 3 |
+
size 4997743184
|
model-00007-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:cd6576efaf9c1b176ef5e955e3e6e6baa6c21a46deda2165c1c71c17350c2b9d
|
| 3 |
+
size 4997743184
|
model-00008-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b956a0b374e5c075dc7396f9ea8225d47183f8385d7fede433f1d7324ff54224
|
| 3 |
+
size 4997743184
|
model-00009-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eb01238b8aece60d17b0597d1328f132c76270bf29fc1083c5049f7c2f3a7197
|
| 3 |
+
size 4997743184
|
model-00010-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:eeaff5cc7406103d016f4cbac73a406ceddd0621f70284783b1bf906a29cd075
|
| 3 |
+
size 4997743184
|
model-00011-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:fe772ef9d9d400866bbc5459a3af83cae5df485d924eb0bc05e6c56f64dab2e1
|
| 3 |
+
size 4997743184
|
model-00012-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c300d2dbef4e7ecec3e272d4cfa9366716b06c1e96b3ccaebd4532aa73b8ac95
|
| 3 |
+
size 4997743184
|
model-00013-of-00013.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:db3f3a3ee279104560a921bc242900266fd3cdc98a6d6d4958dd278765350eb5
|
| 3 |
+
size 1094220288
|
model.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|im_start|>",
|
| 4 |
+
"<|im_end|>",
|
| 5 |
+
"<|object_ref_start|>",
|
| 6 |
+
"<|object_ref_end|>",
|
| 7 |
+
"<|box_start|>",
|
| 8 |
+
"<|box_end|>",
|
| 9 |
+
"<|quad_start|>",
|
| 10 |
+
"<|quad_end|>",
|
| 11 |
+
"<|vision_start|>",
|
| 12 |
+
"<|vision_end|>",
|
| 13 |
+
"<|vision_pad|>",
|
| 14 |
+
"<|image_pad|>",
|
| 15 |
+
"<|video_pad|>"
|
| 16 |
+
],
|
| 17 |
+
"eos_token": {
|
| 18 |
+
"content": "<|im_end|>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
},
|
| 24 |
+
"pad_token": {
|
| 25 |
+
"content": "<|endoftext|>",
|
| 26 |
+
"lstrip": false,
|
| 27 |
+
"normalized": false,
|
| 28 |
+
"rstrip": false,
|
| 29 |
+
"single_word": false
|
| 30 |
+
}
|
| 31 |
+
}
|
tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
|
| 3 |
+
size 11422654
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,241 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"151643": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"151644": {
|
| 14 |
+
"content": "<|im_start|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"151645": {
|
| 22 |
+
"content": "<|im_end|>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
},
|
| 29 |
+
"151646": {
|
| 30 |
+
"content": "<|object_ref_start|>",
|
| 31 |
+
"lstrip": false,
|
| 32 |
+
"normalized": false,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"special": true
|
| 36 |
+
},
|
| 37 |
+
"151647": {
|
| 38 |
+
"content": "<|object_ref_end|>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false,
|
| 43 |
+
"special": true
|
| 44 |
+
},
|
| 45 |
+
"151648": {
|
| 46 |
+
"content": "<|box_start|>",
|
| 47 |
+
"lstrip": false,
|
| 48 |
+
"normalized": false,
|
| 49 |
+
"rstrip": false,
|
| 50 |
+
"single_word": false,
|
| 51 |
+
"special": true
|
| 52 |
+
},
|
| 53 |
+
"151649": {
|
| 54 |
+
"content": "<|box_end|>",
|
| 55 |
+
"lstrip": false,
|
| 56 |
+
"normalized": false,
|
| 57 |
+
"rstrip": false,
|
| 58 |
+
"single_word": false,
|
| 59 |
+
"special": true
|
| 60 |
+
},
|
| 61 |
+
"151650": {
|
| 62 |
+
"content": "<|quad_start|>",
|
| 63 |
+
"lstrip": false,
|
| 64 |
+
"normalized": false,
|
| 65 |
+
"rstrip": false,
|
| 66 |
+
"single_word": false,
|
| 67 |
+
"special": true
|
| 68 |
+
},
|
| 69 |
+
"151651": {
|
| 70 |
+
"content": "<|quad_end|>",
|
| 71 |
+
"lstrip": false,
|
| 72 |
+
"normalized": false,
|
| 73 |
+
"rstrip": false,
|
| 74 |
+
"single_word": false,
|
| 75 |
+
"special": true
|
| 76 |
+
},
|
| 77 |
+
"151652": {
|
| 78 |
+
"content": "<|vision_start|>",
|
| 79 |
+
"lstrip": false,
|
| 80 |
+
"normalized": false,
|
| 81 |
+
"rstrip": false,
|
| 82 |
+
"single_word": false,
|
| 83 |
+
"special": true
|
| 84 |
+
},
|
| 85 |
+
"151653": {
|
| 86 |
+
"content": "<|vision_end|>",
|
| 87 |
+
"lstrip": false,
|
| 88 |
+
"normalized": false,
|
| 89 |
+
"rstrip": false,
|
| 90 |
+
"single_word": false,
|
| 91 |
+
"special": true
|
| 92 |
+
},
|
| 93 |
+
"151654": {
|
| 94 |
+
"content": "<|vision_pad|>",
|
| 95 |
+
"lstrip": false,
|
| 96 |
+
"normalized": false,
|
| 97 |
+
"rstrip": false,
|
| 98 |
+
"single_word": false,
|
| 99 |
+
"special": true
|
| 100 |
+
},
|
| 101 |
+
"151655": {
|
| 102 |
+
"content": "<|image_pad|>",
|
| 103 |
+
"lstrip": false,
|
| 104 |
+
"normalized": false,
|
| 105 |
+
"rstrip": false,
|
| 106 |
+
"single_word": false,
|
| 107 |
+
"special": true
|
| 108 |
+
},
|
| 109 |
+
"151656": {
|
| 110 |
+
"content": "<|video_pad|>",
|
| 111 |
+
"lstrip": false,
|
| 112 |
+
"normalized": false,
|
| 113 |
+
"rstrip": false,
|
| 114 |
+
"single_word": false,
|
| 115 |
+
"special": true
|
| 116 |
+
},
|
| 117 |
+
"151657": {
|
| 118 |
+
"content": "<tool_call>",
|
| 119 |
+
"lstrip": false,
|
| 120 |
+
"normalized": false,
|
| 121 |
+
"rstrip": false,
|
| 122 |
+
"single_word": false,
|
| 123 |
+
"special": false
|
| 124 |
+
},
|
| 125 |
+
"151658": {
|
| 126 |
+
"content": "</tool_call>",
|
| 127 |
+
"lstrip": false,
|
| 128 |
+
"normalized": false,
|
| 129 |
+
"rstrip": false,
|
| 130 |
+
"single_word": false,
|
| 131 |
+
"special": false
|
| 132 |
+
},
|
| 133 |
+
"151659": {
|
| 134 |
+
"content": "<|fim_prefix|>",
|
| 135 |
+
"lstrip": false,
|
| 136 |
+
"normalized": false,
|
| 137 |
+
"rstrip": false,
|
| 138 |
+
"single_word": false,
|
| 139 |
+
"special": false
|
| 140 |
+
},
|
| 141 |
+
"151660": {
|
| 142 |
+
"content": "<|fim_middle|>",
|
| 143 |
+
"lstrip": false,
|
| 144 |
+
"normalized": false,
|
| 145 |
+
"rstrip": false,
|
| 146 |
+
"single_word": false,
|
| 147 |
+
"special": false
|
| 148 |
+
},
|
| 149 |
+
"151661": {
|
| 150 |
+
"content": "<|fim_suffix|>",
|
| 151 |
+
"lstrip": false,
|
| 152 |
+
"normalized": false,
|
| 153 |
+
"rstrip": false,
|
| 154 |
+
"single_word": false,
|
| 155 |
+
"special": false
|
| 156 |
+
},
|
| 157 |
+
"151662": {
|
| 158 |
+
"content": "<|fim_pad|>",
|
| 159 |
+
"lstrip": false,
|
| 160 |
+
"normalized": false,
|
| 161 |
+
"rstrip": false,
|
| 162 |
+
"single_word": false,
|
| 163 |
+
"special": false
|
| 164 |
+
},
|
| 165 |
+
"151663": {
|
| 166 |
+
"content": "<|repo_name|>",
|
| 167 |
+
"lstrip": false,
|
| 168 |
+
"normalized": false,
|
| 169 |
+
"rstrip": false,
|
| 170 |
+
"single_word": false,
|
| 171 |
+
"special": false
|
| 172 |
+
},
|
| 173 |
+
"151664": {
|
| 174 |
+
"content": "<|file_sep|>",
|
| 175 |
+
"lstrip": false,
|
| 176 |
+
"normalized": false,
|
| 177 |
+
"rstrip": false,
|
| 178 |
+
"single_word": false,
|
| 179 |
+
"special": false
|
| 180 |
+
},
|
| 181 |
+
"151665": {
|
| 182 |
+
"content": "<tool_response>",
|
| 183 |
+
"lstrip": false,
|
| 184 |
+
"normalized": false,
|
| 185 |
+
"rstrip": false,
|
| 186 |
+
"single_word": false,
|
| 187 |
+
"special": false
|
| 188 |
+
},
|
| 189 |
+
"151666": {
|
| 190 |
+
"content": "</tool_response>",
|
| 191 |
+
"lstrip": false,
|
| 192 |
+
"normalized": false,
|
| 193 |
+
"rstrip": false,
|
| 194 |
+
"single_word": false,
|
| 195 |
+
"special": false
|
| 196 |
+
},
|
| 197 |
+
"151667": {
|
| 198 |
+
"content": "<think>",
|
| 199 |
+
"lstrip": false,
|
| 200 |
+
"normalized": false,
|
| 201 |
+
"rstrip": false,
|
| 202 |
+
"single_word": false,
|
| 203 |
+
"special": false
|
| 204 |
+
},
|
| 205 |
+
"151668": {
|
| 206 |
+
"content": "</think>",
|
| 207 |
+
"lstrip": false,
|
| 208 |
+
"normalized": false,
|
| 209 |
+
"rstrip": false,
|
| 210 |
+
"single_word": false,
|
| 211 |
+
"special": false
|
| 212 |
+
}
|
| 213 |
+
},
|
| 214 |
+
"additional_special_tokens": [
|
| 215 |
+
"<|im_start|>",
|
| 216 |
+
"<|im_end|>",
|
| 217 |
+
"<|object_ref_start|>",
|
| 218 |
+
"<|object_ref_end|>",
|
| 219 |
+
"<|box_start|>",
|
| 220 |
+
"<|box_end|>",
|
| 221 |
+
"<|quad_start|>",
|
| 222 |
+
"<|quad_end|>",
|
| 223 |
+
"<|vision_start|>",
|
| 224 |
+
"<|vision_end|>",
|
| 225 |
+
"<|vision_pad|>",
|
| 226 |
+
"<|image_pad|>",
|
| 227 |
+
"<|video_pad|>"
|
| 228 |
+
],
|
| 229 |
+
"bos_token": null,
|
| 230 |
+
"chat_template": "{%- if tools %}\n {{- '<|im_start|>system\\n' }}\n {%- if messages[0].role == 'system' %}\n {{- messages[0].content + '\\n\\n' }}\n {%- endif %}\n {{- \"# Tools\\n\\nYou may call one or more functions to assist with the user query.\\n\\nYou are provided with function signatures within <tools></tools> XML tags:\\n<tools>\" }}\n {%- for tool in tools %}\n {{- \"\\n\" }}\n {{- tool | tojson }}\n {%- endfor %}\n {{- \"\\n</tools>\\n\\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\\n<tool_call>\\n{\\\"name\\\": <function-name>, \\\"arguments\\\": <args-json-object>}\\n</tool_call><|im_end|>\\n\" }}\n{%- else %}\n {%- if messages[0].role == 'system' %}\n {{- '<|im_start|>system\\n' + messages[0].content + '<|im_end|>\\n' }}\n {%- endif %}\n{%- endif %}\n{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}\n{%- for message in messages[::-1] %}\n {%- set index = (messages|length - 1) - loop.index0 %}\n {%- if ns.multi_step_tool and message.role == \"user\" and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}\n {%- set ns.multi_step_tool = false %}\n {%- set ns.last_query_index = index %}\n {%- endif %}\n{%- endfor %}\n{%- for message in messages %}\n {%- if (message.role == \"user\") or (message.role == \"system\" and not loop.first) %}\n {{- '<|im_start|>' + message.role + '\\n' + message.content + '<|im_end|>' + '\\n' }}\n {%- elif message.role == \"assistant\" %}\n {%- set content = message.content %}\n {%- set reasoning_content = '' %}\n {%- if message.reasoning_content is defined and message.reasoning_content is not none %}\n {%- set reasoning_content = message.reasoning_content %}\n {%- else %}\n {%- if '</think>' in message.content %}\n {%- set content = message.content.split('</think>')[-1].lstrip('\\n') %}\n {%- set reasoning_content = message.content.split('</think>')[0].rstrip('\\n').split('<think>')[-1].lstrip('\\n') %}\n {%- endif %}\n {%- endif %}\n {%- if loop.index0 > ns.last_query_index %}\n {%- if loop.last or (not loop.last and reasoning_content) %}\n {{- '<|im_start|>' + message.role + '\\n<think>\\n' + reasoning_content.strip('\\n') + '\\n</think>\\n\\n' + content.lstrip('\\n') }}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- else %}\n {{- '<|im_start|>' + message.role + '\\n' + content }}\n {%- endif %}\n {%- if message.tool_calls %}\n {%- for tool_call in message.tool_calls %}\n {%- if (loop.first and content) or (not loop.first) %}\n {{- '\\n' }}\n {%- endif %}\n {%- if tool_call.function %}\n {%- set tool_call = tool_call.function %}\n {%- endif %}\n {{- '<tool_call>\\n{\"name\": \"' }}\n {{- tool_call.name }}\n {{- '\", \"arguments\": ' }}\n {%- if tool_call.arguments is string %}\n {{- tool_call.arguments }}\n {%- else %}\n {{- tool_call.arguments | tojson }}\n {%- endif %}\n {{- '}\\n</tool_call>' }}\n {%- endfor %}\n {%- endif %}\n {{- '<|im_end|>\\n' }}\n {%- elif message.role == \"tool\" %}\n {%- if loop.first or (messages[loop.index0 - 1].role != \"tool\") %}\n {{- '<|im_start|>user' }}\n {%- endif %}\n {{- '\\n<tool_response>\\n' }}\n {{- message.content }}\n {{- '\\n</tool_response>' }}\n {%- if loop.last or (messages[loop.index0 + 1].role != \"tool\") %}\n {{- '<|im_end|>\\n' }}\n {%- endif %}\n {%- endif %}\n{%- endfor %}\n{%- if add_generation_prompt %}\n {{- '<|im_start|>assistant\\n' }}\n {%- if enable_thinking is defined and enable_thinking is false %}\n {{- '<think>\\n\\n</think>\\n\\n' }}\n {%- endif %}\n{%- endif %}",
|
| 231 |
+
"clean_up_tokenization_spaces": false,
|
| 232 |
+
"eos_token": "<|im_end|>",
|
| 233 |
+
"errors": "replace",
|
| 234 |
+
"extra_special_tokens": {},
|
| 235 |
+
"model_max_length": 131072,
|
| 236 |
+
"pad_token": "<|endoftext|>",
|
| 237 |
+
"padding_side": "right",
|
| 238 |
+
"split_special_tokens": false,
|
| 239 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 240 |
+
"unk_token": null
|
| 241 |
+
}
|
train_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"epoch": 5.0,
|
| 3 |
+
"total_flos": 78480301031424.0,
|
| 4 |
+
"train_loss": 0.7362074042123462,
|
| 5 |
+
"train_runtime": 9541.1641,
|
| 6 |
+
"train_samples_per_second": 0.524,
|
| 7 |
+
"train_steps_per_second": 0.033
|
| 8 |
+
}
|
trainer_log.jsonl
ADDED
|
@@ -0,0 +1,316 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{"current_steps": 1, "total_steps": 315, "loss": 1.0469, "lr": 0.0, "epoch": 0.015873015873015872, "percentage": 0.32, "elapsed_time": "0:00:38", "remaining_time": "3:19:34"}
|
| 2 |
+
{"current_steps": 2, "total_steps": 315, "loss": 0.9929, "lr": 3.125e-07, "epoch": 0.031746031746031744, "percentage": 0.63, "elapsed_time": "0:01:08", "remaining_time": "2:58:51"}
|
| 3 |
+
{"current_steps": 3, "total_steps": 315, "loss": 1.0189, "lr": 6.25e-07, "epoch": 0.047619047619047616, "percentage": 0.95, "elapsed_time": "0:01:35", "remaining_time": "2:46:13"}
|
| 4 |
+
{"current_steps": 4, "total_steps": 315, "loss": 0.9098, "lr": 9.375000000000001e-07, "epoch": 0.06349206349206349, "percentage": 1.27, "elapsed_time": "0:02:04", "remaining_time": "2:41:31"}
|
| 5 |
+
{"current_steps": 5, "total_steps": 315, "loss": 1.0462, "lr": 1.25e-06, "epoch": 0.07936507936507936, "percentage": 1.59, "elapsed_time": "0:02:33", "remaining_time": "2:38:40"}
|
| 6 |
+
{"current_steps": 6, "total_steps": 315, "loss": 0.9986, "lr": 1.5625e-06, "epoch": 0.09523809523809523, "percentage": 1.9, "elapsed_time": "0:03:02", "remaining_time": "2:36:32"}
|
| 7 |
+
{"current_steps": 7, "total_steps": 315, "loss": 0.9554, "lr": 1.8750000000000003e-06, "epoch": 0.1111111111111111, "percentage": 2.22, "elapsed_time": "0:03:30", "remaining_time": "2:34:36"}
|
| 8 |
+
{"current_steps": 8, "total_steps": 315, "loss": 0.9522, "lr": 2.1875000000000002e-06, "epoch": 0.12698412698412698, "percentage": 2.54, "elapsed_time": "0:03:59", "remaining_time": "2:33:10"}
|
| 9 |
+
{"current_steps": 9, "total_steps": 315, "loss": 0.9994, "lr": 2.5e-06, "epoch": 0.14285714285714285, "percentage": 2.86, "elapsed_time": "0:04:28", "remaining_time": "2:31:53"}
|
| 10 |
+
{"current_steps": 10, "total_steps": 315, "loss": 0.9373, "lr": 2.8125e-06, "epoch": 0.15873015873015872, "percentage": 3.17, "elapsed_time": "0:04:56", "remaining_time": "2:30:43"}
|
| 11 |
+
{"current_steps": 11, "total_steps": 315, "loss": 0.8839, "lr": 3.125e-06, "epoch": 0.1746031746031746, "percentage": 3.49, "elapsed_time": "0:05:25", "remaining_time": "2:29:55"}
|
| 12 |
+
{"current_steps": 12, "total_steps": 315, "loss": 0.9907, "lr": 3.4375e-06, "epoch": 0.19047619047619047, "percentage": 3.81, "elapsed_time": "0:05:53", "remaining_time": "2:28:56"}
|
| 13 |
+
{"current_steps": 13, "total_steps": 315, "loss": 0.9322, "lr": 3.7500000000000005e-06, "epoch": 0.20634920634920634, "percentage": 4.13, "elapsed_time": "0:06:20", "remaining_time": "2:27:10"}
|
| 14 |
+
{"current_steps": 14, "total_steps": 315, "loss": 1.0934, "lr": 4.0625000000000005e-06, "epoch": 0.2222222222222222, "percentage": 4.44, "elapsed_time": "0:06:48", "remaining_time": "2:26:28"}
|
| 15 |
+
{"current_steps": 15, "total_steps": 315, "loss": 0.9771, "lr": 4.3750000000000005e-06, "epoch": 0.23809523809523808, "percentage": 4.76, "elapsed_time": "0:07:17", "remaining_time": "2:25:47"}
|
| 16 |
+
{"current_steps": 16, "total_steps": 315, "loss": 1.0442, "lr": 4.6875000000000004e-06, "epoch": 0.25396825396825395, "percentage": 5.08, "elapsed_time": "0:07:47", "remaining_time": "2:25:31"}
|
| 17 |
+
{"current_steps": 17, "total_steps": 315, "loss": 0.949, "lr": 5e-06, "epoch": 0.2698412698412698, "percentage": 5.4, "elapsed_time": "0:08:15", "remaining_time": "2:24:50"}
|
| 18 |
+
{"current_steps": 18, "total_steps": 315, "loss": 1.0098, "lr": 5.3125e-06, "epoch": 0.2857142857142857, "percentage": 5.71, "elapsed_time": "0:08:44", "remaining_time": "2:24:13"}
|
| 19 |
+
{"current_steps": 19, "total_steps": 315, "loss": 0.7468, "lr": 5.625e-06, "epoch": 0.30158730158730157, "percentage": 6.03, "elapsed_time": "0:09:05", "remaining_time": "2:21:41"}
|
| 20 |
+
{"current_steps": 20, "total_steps": 315, "loss": 0.9232, "lr": 5.9375e-06, "epoch": 0.31746031746031744, "percentage": 6.35, "elapsed_time": "0:09:35", "remaining_time": "2:21:21"}
|
| 21 |
+
{"current_steps": 21, "total_steps": 315, "loss": 0.9885, "lr": 6.25e-06, "epoch": 0.3333333333333333, "percentage": 6.67, "elapsed_time": "0:10:03", "remaining_time": "2:20:55"}
|
| 22 |
+
{"current_steps": 22, "total_steps": 315, "loss": 1.0508, "lr": 6.5625e-06, "epoch": 0.3492063492063492, "percentage": 6.98, "elapsed_time": "0:10:33", "remaining_time": "2:20:38"}
|
| 23 |
+
{"current_steps": 23, "total_steps": 315, "loss": 0.8728, "lr": 6.875e-06, "epoch": 0.36507936507936506, "percentage": 7.3, "elapsed_time": "0:11:02", "remaining_time": "2:20:16"}
|
| 24 |
+
{"current_steps": 24, "total_steps": 315, "loss": 0.8349, "lr": 7.1875e-06, "epoch": 0.38095238095238093, "percentage": 7.62, "elapsed_time": "0:11:31", "remaining_time": "2:19:44"}
|
| 25 |
+
{"current_steps": 25, "total_steps": 315, "loss": 1.0283, "lr": 7.500000000000001e-06, "epoch": 0.3968253968253968, "percentage": 7.94, "elapsed_time": "0:11:59", "remaining_time": "2:19:02"}
|
| 26 |
+
{"current_steps": 26, "total_steps": 315, "loss": 0.8612, "lr": 7.8125e-06, "epoch": 0.4126984126984127, "percentage": 8.25, "elapsed_time": "0:12:27", "remaining_time": "2:18:31"}
|
| 27 |
+
{"current_steps": 27, "total_steps": 315, "loss": 0.7232, "lr": 8.125000000000001e-06, "epoch": 0.42857142857142855, "percentage": 8.57, "elapsed_time": "0:12:56", "remaining_time": "2:17:59"}
|
| 28 |
+
{"current_steps": 28, "total_steps": 315, "loss": 0.9752, "lr": 8.4375e-06, "epoch": 0.4444444444444444, "percentage": 8.89, "elapsed_time": "0:13:24", "remaining_time": "2:17:30"}
|
| 29 |
+
{"current_steps": 29, "total_steps": 315, "loss": 0.9053, "lr": 8.750000000000001e-06, "epoch": 0.4603174603174603, "percentage": 9.21, "elapsed_time": "0:13:53", "remaining_time": "2:17:00"}
|
| 30 |
+
{"current_steps": 30, "total_steps": 315, "loss": 0.9479, "lr": 9.0625e-06, "epoch": 0.47619047619047616, "percentage": 9.52, "elapsed_time": "0:14:22", "remaining_time": "2:16:29"}
|
| 31 |
+
{"current_steps": 31, "total_steps": 315, "loss": 0.9188, "lr": 9.375000000000001e-06, "epoch": 0.49206349206349204, "percentage": 9.84, "elapsed_time": "0:14:50", "remaining_time": "2:15:59"}
|
| 32 |
+
{"current_steps": 32, "total_steps": 315, "loss": 0.8847, "lr": 9.6875e-06, "epoch": 0.5079365079365079, "percentage": 10.16, "elapsed_time": "0:15:17", "remaining_time": "2:15:15"}
|
| 33 |
+
{"current_steps": 33, "total_steps": 315, "loss": 1.0379, "lr": 1e-05, "epoch": 0.5238095238095238, "percentage": 10.48, "elapsed_time": "0:15:46", "remaining_time": "2:14:45"}
|
| 34 |
+
{"current_steps": 34, "total_steps": 315, "loss": 0.8376, "lr": 9.999691920767945e-06, "epoch": 0.5396825396825397, "percentage": 10.79, "elapsed_time": "0:16:14", "remaining_time": "2:14:16"}
|
| 35 |
+
{"current_steps": 35, "total_steps": 315, "loss": 0.8241, "lr": 9.998767721036901e-06, "epoch": 0.5555555555555556, "percentage": 11.11, "elapsed_time": "0:16:43", "remaining_time": "2:13:48"}
|
| 36 |
+
{"current_steps": 36, "total_steps": 315, "loss": 0.9693, "lr": 9.997227514697568e-06, "epoch": 0.5714285714285714, "percentage": 11.43, "elapsed_time": "0:17:12", "remaining_time": "2:13:21"}
|
| 37 |
+
{"current_steps": 37, "total_steps": 315, "loss": 0.9839, "lr": 9.99507149155218e-06, "epoch": 0.5873015873015873, "percentage": 11.75, "elapsed_time": "0:17:41", "remaining_time": "2:12:52"}
|
| 38 |
+
{"current_steps": 38, "total_steps": 315, "loss": 0.8479, "lr": 9.992299917291118e-06, "epoch": 0.6031746031746031, "percentage": 12.06, "elapsed_time": "0:18:09", "remaining_time": "2:12:23"}
|
| 39 |
+
{"current_steps": 39, "total_steps": 315, "loss": 0.9095, "lr": 9.98891313346017e-06, "epoch": 0.6190476190476191, "percentage": 12.38, "elapsed_time": "0:18:38", "remaining_time": "2:11:57"}
|
| 40 |
+
{"current_steps": 40, "total_steps": 315, "loss": 0.7685, "lr": 9.984911557418444e-06, "epoch": 0.6349206349206349, "percentage": 12.7, "elapsed_time": "0:19:07", "remaining_time": "2:11:28"}
|
| 41 |
+
{"current_steps": 41, "total_steps": 315, "loss": 0.8387, "lr": 9.980295682286924e-06, "epoch": 0.6507936507936508, "percentage": 13.02, "elapsed_time": "0:19:35", "remaining_time": "2:10:58"}
|
| 42 |
+
{"current_steps": 42, "total_steps": 315, "loss": 0.9107, "lr": 9.97506607688772e-06, "epoch": 0.6666666666666666, "percentage": 13.33, "elapsed_time": "0:20:05", "remaining_time": "2:10:32"}
|
| 43 |
+
{"current_steps": 43, "total_steps": 315, "loss": 0.8308, "lr": 9.969223385673958e-06, "epoch": 0.6825396825396826, "percentage": 13.65, "elapsed_time": "0:20:34", "remaining_time": "2:10:06"}
|
| 44 |
+
{"current_steps": 44, "total_steps": 315, "loss": 0.7516, "lr": 9.962768328650367e-06, "epoch": 0.6984126984126984, "percentage": 13.97, "elapsed_time": "0:21:02", "remaining_time": "2:09:37"}
|
| 45 |
+
{"current_steps": 45, "total_steps": 315, "loss": 0.8443, "lr": 9.95570170128455e-06, "epoch": 0.7142857142857143, "percentage": 14.29, "elapsed_time": "0:21:31", "remaining_time": "2:09:09"}
|
| 46 |
+
{"current_steps": 46, "total_steps": 315, "loss": 0.7959, "lr": 9.94802437440896e-06, "epoch": 0.7301587301587301, "percentage": 14.6, "elapsed_time": "0:22:00", "remaining_time": "2:08:41"}
|
| 47 |
+
{"current_steps": 47, "total_steps": 315, "loss": 0.8964, "lr": 9.939737294113585e-06, "epoch": 0.746031746031746, "percentage": 14.92, "elapsed_time": "0:22:29", "remaining_time": "2:08:13"}
|
| 48 |
+
{"current_steps": 48, "total_steps": 315, "loss": 0.8884, "lr": 9.930841481629358e-06, "epoch": 0.7619047619047619, "percentage": 15.24, "elapsed_time": "0:22:57", "remaining_time": "2:07:43"}
|
| 49 |
+
{"current_steps": 49, "total_steps": 315, "loss": 0.7817, "lr": 9.92133803320231e-06, "epoch": 0.7777777777777778, "percentage": 15.56, "elapsed_time": "0:23:26", "remaining_time": "2:07:13"}
|
| 50 |
+
{"current_steps": 50, "total_steps": 315, "loss": 0.819, "lr": 9.91122811995848e-06, "epoch": 0.7936507936507936, "percentage": 15.87, "elapsed_time": "0:23:51", "remaining_time": "2:06:25"}
|
| 51 |
+
{"current_steps": 51, "total_steps": 315, "loss": 0.8691, "lr": 9.90051298775959e-06, "epoch": 0.8095238095238095, "percentage": 16.19, "elapsed_time": "0:24:19", "remaining_time": "2:05:55"}
|
| 52 |
+
{"current_steps": 52, "total_steps": 315, "loss": 0.8259, "lr": 9.88919395704952e-06, "epoch": 0.8253968253968254, "percentage": 16.51, "elapsed_time": "0:24:48", "remaining_time": "2:05:28"}
|
| 53 |
+
{"current_steps": 53, "total_steps": 315, "loss": 0.9318, "lr": 9.877272422691583e-06, "epoch": 0.8412698412698413, "percentage": 16.83, "elapsed_time": "0:25:17", "remaining_time": "2:04:59"}
|
| 54 |
+
{"current_steps": 54, "total_steps": 315, "loss": 0.7983, "lr": 9.864749853796642e-06, "epoch": 0.8571428571428571, "percentage": 17.14, "elapsed_time": "0:25:45", "remaining_time": "2:04:31"}
|
| 55 |
+
{"current_steps": 55, "total_steps": 315, "loss": 0.7289, "lr": 9.85162779354206e-06, "epoch": 0.873015873015873, "percentage": 17.46, "elapsed_time": "0:26:14", "remaining_time": "2:04:02"}
|
| 56 |
+
{"current_steps": 56, "total_steps": 315, "loss": 0.8795, "lr": 9.837907858981536e-06, "epoch": 0.8888888888888888, "percentage": 17.78, "elapsed_time": "0:26:43", "remaining_time": "2:03:34"}
|
| 57 |
+
{"current_steps": 57, "total_steps": 315, "loss": 0.8625, "lr": 9.823591740845831e-06, "epoch": 0.9047619047619048, "percentage": 18.1, "elapsed_time": "0:27:11", "remaining_time": "2:03:06"}
|
| 58 |
+
{"current_steps": 58, "total_steps": 315, "loss": 0.7975, "lr": 9.808681203334416e-06, "epoch": 0.9206349206349206, "percentage": 18.41, "elapsed_time": "0:27:40", "remaining_time": "2:02:36"}
|
| 59 |
+
{"current_steps": 59, "total_steps": 315, "loss": 0.878, "lr": 9.793178083898073e-06, "epoch": 0.9365079365079365, "percentage": 18.73, "elapsed_time": "0:28:08", "remaining_time": "2:02:07"}
|
| 60 |
+
{"current_steps": 60, "total_steps": 315, "loss": 0.842, "lr": 9.777084293012448e-06, "epoch": 0.9523809523809523, "percentage": 19.05, "elapsed_time": "0:28:37", "remaining_time": "2:01:38"}
|
| 61 |
+
{"current_steps": 61, "total_steps": 315, "loss": 0.7661, "lr": 9.760401813942641e-06, "epoch": 0.9682539682539683, "percentage": 19.37, "elapsed_time": "0:29:05", "remaining_time": "2:01:09"}
|
| 62 |
+
{"current_steps": 62, "total_steps": 315, "loss": 0.8685, "lr": 9.743132702498785e-06, "epoch": 0.9841269841269841, "percentage": 19.68, "elapsed_time": "0:29:34", "remaining_time": "2:00:40"}
|
| 63 |
+
{"current_steps": 63, "total_steps": 315, "loss": 0.7676, "lr": 9.725279086782719e-06, "epoch": 1.0, "percentage": 20.0, "elapsed_time": "0:30:03", "remaining_time": "2:00:12"}
|
| 64 |
+
{"current_steps": 64, "total_steps": 315, "loss": 0.7978, "lr": 9.706843166925733e-06, "epoch": 1.0158730158730158, "percentage": 20.32, "elapsed_time": "0:32:18", "remaining_time": "2:06:42"}
|
| 65 |
+
{"current_steps": 65, "total_steps": 315, "loss": 0.8264, "lr": 9.687827214817433e-06, "epoch": 1.0317460317460316, "percentage": 20.63, "elapsed_time": "0:32:40", "remaining_time": "2:05:40"}
|
| 66 |
+
{"current_steps": 66, "total_steps": 315, "loss": 0.8898, "lr": 9.668233573825794e-06, "epoch": 1.0476190476190477, "percentage": 20.95, "elapsed_time": "0:33:07", "remaining_time": "2:04:59"}
|
| 67 |
+
{"current_steps": 67, "total_steps": 315, "loss": 0.7317, "lr": 9.64806465850836e-06, "epoch": 1.0634920634920635, "percentage": 21.27, "elapsed_time": "0:33:36", "remaining_time": "2:04:23"}
|
| 68 |
+
{"current_steps": 68, "total_steps": 315, "loss": 0.7307, "lr": 9.62732295431471e-06, "epoch": 1.0793650793650793, "percentage": 21.59, "elapsed_time": "0:34:04", "remaining_time": "2:03:47"}
|
| 69 |
+
{"current_steps": 69, "total_steps": 315, "loss": 0.8977, "lr": 9.606011017280166e-06, "epoch": 1.0952380952380953, "percentage": 21.9, "elapsed_time": "0:34:33", "remaining_time": "2:03:12"}
|
| 70 |
+
{"current_steps": 70, "total_steps": 315, "loss": 0.7648, "lr": 9.5841314737108e-06, "epoch": 1.1111111111111112, "percentage": 22.22, "elapsed_time": "0:35:02", "remaining_time": "2:02:37"}
|
| 71 |
+
{"current_steps": 71, "total_steps": 315, "loss": 0.7995, "lr": 9.56168701985981e-06, "epoch": 1.126984126984127, "percentage": 22.54, "elapsed_time": "0:35:31", "remaining_time": "2:02:04"}
|
| 72 |
+
{"current_steps": 72, "total_steps": 315, "loss": 0.8072, "lr": 9.538680421595236e-06, "epoch": 1.1428571428571428, "percentage": 22.86, "elapsed_time": "0:36:00", "remaining_time": "2:01:31"}
|
| 73 |
+
{"current_steps": 73, "total_steps": 315, "loss": 0.8128, "lr": 9.515114514059127e-06, "epoch": 1.1587301587301586, "percentage": 23.17, "elapsed_time": "0:36:29", "remaining_time": "2:00:57"}
|
| 74 |
+
{"current_steps": 74, "total_steps": 315, "loss": 0.7876, "lr": 9.490992201318165e-06, "epoch": 1.1746031746031746, "percentage": 23.49, "elapsed_time": "0:36:58", "remaining_time": "2:00:23"}
|
| 75 |
+
{"current_steps": 75, "total_steps": 315, "loss": 0.7755, "lr": 9.466316456005783e-06, "epoch": 1.1904761904761905, "percentage": 23.81, "elapsed_time": "0:37:26", "remaining_time": "1:59:49"}
|
| 76 |
+
{"current_steps": 76, "total_steps": 315, "loss": 0.7015, "lr": 9.441090318955843e-06, "epoch": 1.2063492063492063, "percentage": 24.13, "elapsed_time": "0:37:56", "remaining_time": "1:59:17"}
|
| 77 |
+
{"current_steps": 77, "total_steps": 315, "loss": 0.7346, "lr": 9.415316898827923e-06, "epoch": 1.2222222222222223, "percentage": 24.44, "elapsed_time": "0:38:25", "remaining_time": "1:58:44"}
|
| 78 |
+
{"current_steps": 78, "total_steps": 315, "loss": 0.8242, "lr": 9.388999371724212e-06, "epoch": 1.2380952380952381, "percentage": 24.76, "elapsed_time": "0:38:53", "remaining_time": "1:58:10"}
|
| 79 |
+
{"current_steps": 79, "total_steps": 315, "loss": 0.8928, "lr": 9.362140980798127e-06, "epoch": 1.253968253968254, "percentage": 25.08, "elapsed_time": "0:39:21", "remaining_time": "1:57:33"}
|
| 80 |
+
{"current_steps": 80, "total_steps": 315, "loss": 0.7581, "lr": 9.334745035854646e-06, "epoch": 1.2698412698412698, "percentage": 25.4, "elapsed_time": "0:39:49", "remaining_time": "1:56:59"}
|
| 81 |
+
{"current_steps": 81, "total_steps": 315, "loss": 0.8361, "lr": 9.306814912942445e-06, "epoch": 1.2857142857142856, "percentage": 25.71, "elapsed_time": "0:40:18", "remaining_time": "1:56:26"}
|
| 82 |
+
{"current_steps": 82, "total_steps": 315, "loss": 0.7794, "lr": 9.278354053937848e-06, "epoch": 1.3015873015873016, "percentage": 26.03, "elapsed_time": "0:40:46", "remaining_time": "1:55:53"}
|
| 83 |
+
{"current_steps": 83, "total_steps": 315, "loss": 0.8542, "lr": 9.249365966120692e-06, "epoch": 1.3174603174603174, "percentage": 26.35, "elapsed_time": "0:41:15", "remaining_time": "1:55:19"}
|
| 84 |
+
{"current_steps": 84, "total_steps": 315, "loss": 0.8101, "lr": 9.219854221742106e-06, "epoch": 1.3333333333333333, "percentage": 26.67, "elapsed_time": "0:41:44", "remaining_time": "1:54:48"}
|
| 85 |
+
{"current_steps": 85, "total_steps": 315, "loss": 0.7419, "lr": 9.189822457584311e-06, "epoch": 1.3492063492063493, "percentage": 26.98, "elapsed_time": "0:42:13", "remaining_time": "1:54:14"}
|
| 86 |
+
{"current_steps": 86, "total_steps": 315, "loss": 0.6576, "lr": 9.159274374512444e-06, "epoch": 1.3650793650793651, "percentage": 27.3, "elapsed_time": "0:42:41", "remaining_time": "1:53:41"}
|
| 87 |
+
{"current_steps": 87, "total_steps": 315, "loss": 0.8058, "lr": 9.128213737018493e-06, "epoch": 1.380952380952381, "percentage": 27.62, "elapsed_time": "0:43:10", "remaining_time": "1:53:08"}
|
| 88 |
+
{"current_steps": 88, "total_steps": 315, "loss": 0.8849, "lr": 9.096644372757393e-06, "epoch": 1.3968253968253967, "percentage": 27.94, "elapsed_time": "0:43:39", "remaining_time": "1:52:36"}
|
| 89 |
+
{"current_steps": 89, "total_steps": 315, "loss": 0.7969, "lr": 9.064570172075349e-06, "epoch": 1.4126984126984126, "percentage": 28.25, "elapsed_time": "0:44:07", "remaining_time": "1:52:03"}
|
| 90 |
+
{"current_steps": 90, "total_steps": 315, "loss": 0.7983, "lr": 9.031995087530403e-06, "epoch": 1.4285714285714286, "percentage": 28.57, "elapsed_time": "0:44:35", "remaining_time": "1:51:29"}
|
| 91 |
+
{"current_steps": 91, "total_steps": 315, "loss": 0.6612, "lr": 8.99892313340537e-06, "epoch": 1.4444444444444444, "percentage": 28.89, "elapsed_time": "0:45:04", "remaining_time": "1:50:56"}
|
| 92 |
+
{"current_steps": 92, "total_steps": 315, "loss": 0.8026, "lr": 8.96535838521314e-06, "epoch": 1.4603174603174602, "percentage": 29.21, "elapsed_time": "0:45:33", "remaining_time": "1:50:24"}
|
| 93 |
+
{"current_steps": 93, "total_steps": 315, "loss": 0.8051, "lr": 8.931304979194452e-06, "epoch": 1.4761904761904763, "percentage": 29.52, "elapsed_time": "0:46:02", "remaining_time": "1:49:53"}
|
| 94 |
+
{"current_steps": 94, "total_steps": 315, "loss": 0.7354, "lr": 8.896767111808177e-06, "epoch": 1.492063492063492, "percentage": 29.84, "elapsed_time": "0:46:30", "remaining_time": "1:49:21"}
|
| 95 |
+
{"current_steps": 95, "total_steps": 315, "loss": 0.9129, "lr": 8.861749039214177e-06, "epoch": 1.507936507936508, "percentage": 30.16, "elapsed_time": "0:46:59", "remaining_time": "1:48:49"}
|
| 96 |
+
{"current_steps": 96, "total_steps": 315, "loss": 0.8445, "lr": 8.826255076748823e-06, "epoch": 1.5238095238095237, "percentage": 30.48, "elapsed_time": "0:47:27", "remaining_time": "1:48:16"}
|
| 97 |
+
{"current_steps": 97, "total_steps": 315, "loss": 0.7212, "lr": 8.790289598393186e-06, "epoch": 1.5396825396825395, "percentage": 30.79, "elapsed_time": "0:47:54", "remaining_time": "1:47:40"}
|
| 98 |
+
{"current_steps": 98, "total_steps": 315, "loss": 0.8149, "lr": 8.753857036234055e-06, "epoch": 1.5555555555555556, "percentage": 31.11, "elapsed_time": "0:48:23", "remaining_time": "1:47:10"}
|
| 99 |
+
{"current_steps": 99, "total_steps": 315, "loss": 0.7365, "lr": 8.716961879917734e-06, "epoch": 1.5714285714285714, "percentage": 31.43, "elapsed_time": "0:48:52", "remaining_time": "1:46:38"}
|
| 100 |
+
{"current_steps": 100, "total_steps": 315, "loss": 0.8131, "lr": 8.679608676096793e-06, "epoch": 1.5873015873015874, "percentage": 31.75, "elapsed_time": "0:49:20", "remaining_time": "1:46:06"}
|
| 101 |
+
{"current_steps": 101, "total_steps": 315, "loss": 0.7946, "lr": 8.641802027869774e-06, "epoch": 1.6031746031746033, "percentage": 32.06, "elapsed_time": "0:49:49", "remaining_time": "1:45:34"}
|
| 102 |
+
{"current_steps": 102, "total_steps": 315, "loss": 0.8574, "lr": 8.603546594213935e-06, "epoch": 1.619047619047619, "percentage": 32.38, "elapsed_time": "0:50:18", "remaining_time": "1:45:02"}
|
| 103 |
+
{"current_steps": 103, "total_steps": 315, "loss": 0.8286, "lr": 8.564847089411128e-06, "epoch": 1.6349206349206349, "percentage": 32.7, "elapsed_time": "0:50:47", "remaining_time": "1:44:33"}
|
| 104 |
+
{"current_steps": 104, "total_steps": 315, "loss": 0.8412, "lr": 8.525708282466839e-06, "epoch": 1.6507936507936507, "percentage": 33.02, "elapsed_time": "0:51:16", "remaining_time": "1:44:02"}
|
| 105 |
+
{"current_steps": 105, "total_steps": 315, "loss": 0.8172, "lr": 8.486134996522502e-06, "epoch": 1.6666666666666665, "percentage": 33.33, "elapsed_time": "0:51:45", "remaining_time": "1:43:31"}
|
| 106 |
+
{"current_steps": 106, "total_steps": 315, "loss": 0.8058, "lr": 8.446132108261136e-06, "epoch": 1.6825396825396826, "percentage": 33.65, "elapsed_time": "0:52:14", "remaining_time": "1:42:59"}
|
| 107 |
+
{"current_steps": 107, "total_steps": 315, "loss": 0.8031, "lr": 8.405704547306379e-06, "epoch": 1.6984126984126984, "percentage": 33.97, "elapsed_time": "0:52:42", "remaining_time": "1:42:28"}
|
| 108 |
+
{"current_steps": 108, "total_steps": 315, "loss": 0.8903, "lr": 8.364857295615006e-06, "epoch": 1.7142857142857144, "percentage": 34.29, "elapsed_time": "0:53:11", "remaining_time": "1:41:57"}
|
| 109 |
+
{"current_steps": 109, "total_steps": 315, "loss": 0.7925, "lr": 8.323595386862985e-06, "epoch": 1.7301587301587302, "percentage": 34.6, "elapsed_time": "0:53:39", "remaining_time": "1:41:25"}
|
| 110 |
+
{"current_steps": 110, "total_steps": 315, "loss": 0.7664, "lr": 8.281923905825188e-06, "epoch": 1.746031746031746, "percentage": 34.92, "elapsed_time": "0:54:06", "remaining_time": "1:40:49"}
|
| 111 |
+
{"current_steps": 111, "total_steps": 315, "loss": 0.7347, "lr": 8.23984798774876e-06, "epoch": 1.7619047619047619, "percentage": 35.24, "elapsed_time": "0:54:34", "remaining_time": "1:40:18"}
|
| 112 |
+
{"current_steps": 112, "total_steps": 315, "loss": 0.7369, "lr": 8.197372817720314e-06, "epoch": 1.7777777777777777, "percentage": 35.56, "elapsed_time": "0:55:03", "remaining_time": "1:39:47"}
|
| 113 |
+
{"current_steps": 113, "total_steps": 315, "loss": 0.7261, "lr": 8.154503630026955e-06, "epoch": 1.7936507936507935, "percentage": 35.87, "elapsed_time": "0:55:32", "remaining_time": "1:39:16"}
|
| 114 |
+
{"current_steps": 114, "total_steps": 315, "loss": 0.7194, "lr": 8.111245707511253e-06, "epoch": 1.8095238095238095, "percentage": 36.19, "elapsed_time": "0:56:00", "remaining_time": "1:38:45"}
|
| 115 |
+
{"current_steps": 115, "total_steps": 315, "loss": 0.7945, "lr": 8.067604380920228e-06, "epoch": 1.8253968253968254, "percentage": 36.51, "elapsed_time": "0:56:29", "remaining_time": "1:38:15"}
|
| 116 |
+
{"current_steps": 116, "total_steps": 315, "loss": 0.8487, "lr": 8.023585028248435e-06, "epoch": 1.8412698412698414, "percentage": 36.83, "elapsed_time": "0:56:58", "remaining_time": "1:37:44"}
|
| 117 |
+
{"current_steps": 117, "total_steps": 315, "loss": 0.8887, "lr": 7.979193074075216e-06, "epoch": 1.8571428571428572, "percentage": 37.14, "elapsed_time": "0:57:27", "remaining_time": "1:37:13"}
|
| 118 |
+
{"current_steps": 118, "total_steps": 315, "loss": 0.6534, "lr": 7.934433988896233e-06, "epoch": 1.873015873015873, "percentage": 37.46, "elapsed_time": "0:57:49", "remaining_time": "1:36:32"}
|
| 119 |
+
{"current_steps": 119, "total_steps": 315, "loss": 0.8214, "lr": 7.889313288449323e-06, "epoch": 1.8888888888888888, "percentage": 37.78, "elapsed_time": "0:58:18", "remaining_time": "1:36:01"}
|
| 120 |
+
{"current_steps": 120, "total_steps": 315, "loss": 0.7614, "lr": 7.843836533034784e-06, "epoch": 1.9047619047619047, "percentage": 38.1, "elapsed_time": "0:58:46", "remaining_time": "1:35:30"}
|
| 121 |
+
{"current_steps": 121, "total_steps": 315, "loss": 0.7996, "lr": 7.798009326830167e-06, "epoch": 1.9206349206349205, "percentage": 38.41, "elapsed_time": "0:59:15", "remaining_time": "1:35:00"}
|
| 122 |
+
{"current_steps": 122, "total_steps": 315, "loss": 0.867, "lr": 7.751837317199673e-06, "epoch": 1.9365079365079365, "percentage": 38.73, "elapsed_time": "0:59:44", "remaining_time": "1:34:30"}
|
| 123 |
+
{"current_steps": 123, "total_steps": 315, "loss": 0.7547, "lr": 7.705326193998207e-06, "epoch": 1.9523809523809523, "percentage": 39.05, "elapsed_time": "1:00:13", "remaining_time": "1:34:01"}
|
| 124 |
+
{"current_steps": 124, "total_steps": 315, "loss": 0.7582, "lr": 7.658481688870218e-06, "epoch": 1.9682539682539684, "percentage": 39.37, "elapsed_time": "1:00:41", "remaining_time": "1:33:28"}
|
| 125 |
+
{"current_steps": 125, "total_steps": 315, "loss": 0.7606, "lr": 7.611309574543373e-06, "epoch": 1.9841269841269842, "percentage": 39.68, "elapsed_time": "1:01:09", "remaining_time": "1:32:58"}
|
| 126 |
+
{"current_steps": 126, "total_steps": 315, "loss": 0.9121, "lr": 7.563815664117173e-06, "epoch": 2.0, "percentage": 40.0, "elapsed_time": "1:01:39", "remaining_time": "1:32:29"}
|
| 127 |
+
{"current_steps": 127, "total_steps": 315, "loss": 0.7122, "lr": 7.5160058103465985e-06, "epoch": 2.015873015873016, "percentage": 40.32, "elapsed_time": "1:03:59", "remaining_time": "1:34:44"}
|
| 128 |
+
{"current_steps": 128, "total_steps": 315, "loss": 0.7567, "lr": 7.467885904920864e-06, "epoch": 2.0317460317460316, "percentage": 40.63, "elapsed_time": "1:04:27", "remaining_time": "1:34:09"}
|
| 129 |
+
{"current_steps": 129, "total_steps": 315, "loss": 0.8318, "lr": 7.419461877737373e-06, "epoch": 2.0476190476190474, "percentage": 40.95, "elapsed_time": "1:04:55", "remaining_time": "1:33:37"}
|
| 130 |
+
{"current_steps": 130, "total_steps": 315, "loss": 0.7428, "lr": 7.370739696170971e-06, "epoch": 2.0634920634920633, "percentage": 41.27, "elapsed_time": "1:05:26", "remaining_time": "1:33:07"}
|
| 131 |
+
{"current_steps": 131, "total_steps": 315, "loss": 0.6161, "lr": 7.321725364338566e-06, "epoch": 2.0793650793650795, "percentage": 41.59, "elapsed_time": "1:05:54", "remaining_time": "1:32:34"}
|
| 132 |
+
{"current_steps": 132, "total_steps": 315, "loss": 0.6432, "lr": 7.272424922359246e-06, "epoch": 2.0952380952380953, "percentage": 41.9, "elapsed_time": "1:06:23", "remaining_time": "1:32:02"}
|
| 133 |
+
{"current_steps": 133, "total_steps": 315, "loss": 0.7817, "lr": 7.222844445609931e-06, "epoch": 2.111111111111111, "percentage": 42.22, "elapsed_time": "1:06:52", "remaining_time": "1:31:30"}
|
| 134 |
+
{"current_steps": 134, "total_steps": 315, "loss": 0.7291, "lr": 7.172990043976703e-06, "epoch": 2.126984126984127, "percentage": 42.54, "elapsed_time": "1:07:21", "remaining_time": "1:30:58"}
|
| 135 |
+
{"current_steps": 135, "total_steps": 315, "loss": 0.7928, "lr": 7.122867861101868e-06, "epoch": 2.142857142857143, "percentage": 42.86, "elapsed_time": "1:07:49", "remaining_time": "1:30:26"}
|
| 136 |
+
{"current_steps": 136, "total_steps": 315, "loss": 0.6864, "lr": 7.072484073626872e-06, "epoch": 2.1587301587301586, "percentage": 43.17, "elapsed_time": "1:08:15", "remaining_time": "1:29:49"}
|
| 137 |
+
{"current_steps": 137, "total_steps": 315, "loss": 0.7627, "lr": 7.021844890431136e-06, "epoch": 2.1746031746031744, "percentage": 43.49, "elapsed_time": "1:08:43", "remaining_time": "1:29:17"}
|
| 138 |
+
{"current_steps": 138, "total_steps": 315, "loss": 0.728, "lr": 6.970956551866925e-06, "epoch": 2.1904761904761907, "percentage": 43.81, "elapsed_time": "1:09:12", "remaining_time": "1:28:45"}
|
| 139 |
+
{"current_steps": 139, "total_steps": 315, "loss": 0.6621, "lr": 6.9198253289903515e-06, "epoch": 2.2063492063492065, "percentage": 44.13, "elapsed_time": "1:09:41", "remaining_time": "1:28:13"}
|
| 140 |
+
{"current_steps": 140, "total_steps": 315, "loss": 0.7351, "lr": 6.868457522788561e-06, "epoch": 2.2222222222222223, "percentage": 44.44, "elapsed_time": "1:10:09", "remaining_time": "1:27:41"}
|
| 141 |
+
{"current_steps": 141, "total_steps": 315, "loss": 0.6568, "lr": 6.816859463403271e-06, "epoch": 2.238095238095238, "percentage": 44.76, "elapsed_time": "1:10:38", "remaining_time": "1:27:10"}
|
| 142 |
+
{"current_steps": 142, "total_steps": 315, "loss": 0.758, "lr": 6.765037509350685e-06, "epoch": 2.253968253968254, "percentage": 45.08, "elapsed_time": "1:11:06", "remaining_time": "1:26:38"}
|
| 143 |
+
{"current_steps": 143, "total_steps": 315, "loss": 0.6657, "lr": 6.7129980467379265e-06, "epoch": 2.2698412698412698, "percentage": 45.4, "elapsed_time": "1:11:33", "remaining_time": "1:26:03"}
|
| 144 |
+
{"current_steps": 144, "total_steps": 315, "loss": 0.6615, "lr": 6.660747488476066e-06, "epoch": 2.2857142857142856, "percentage": 45.71, "elapsed_time": "1:12:02", "remaining_time": "1:25:33"}
|
| 145 |
+
{"current_steps": 145, "total_steps": 315, "loss": 0.6112, "lr": 6.608292273489851e-06, "epoch": 2.3015873015873014, "percentage": 46.03, "elapsed_time": "1:12:31", "remaining_time": "1:25:01"}
|
| 146 |
+
{"current_steps": 146, "total_steps": 315, "loss": 0.7033, "lr": 6.555638865924221e-06, "epoch": 2.317460317460317, "percentage": 46.35, "elapsed_time": "1:12:59", "remaining_time": "1:24:29"}
|
| 147 |
+
{"current_steps": 147, "total_steps": 315, "loss": 0.7578, "lr": 6.502793754347721e-06, "epoch": 2.3333333333333335, "percentage": 46.67, "elapsed_time": "1:13:28", "remaining_time": "1:23:58"}
|
| 148 |
+
{"current_steps": 148, "total_steps": 315, "loss": 0.6863, "lr": 6.449763450952912e-06, "epoch": 2.3492063492063493, "percentage": 46.98, "elapsed_time": "1:13:57", "remaining_time": "1:23:26"}
|
| 149 |
+
{"current_steps": 149, "total_steps": 315, "loss": 0.6825, "lr": 6.396554490753848e-06, "epoch": 2.365079365079365, "percentage": 47.3, "elapsed_time": "1:14:26", "remaining_time": "1:22:55"}
|
| 150 |
+
{"current_steps": 150, "total_steps": 315, "loss": 0.836, "lr": 6.343173430780769e-06, "epoch": 2.380952380952381, "percentage": 47.62, "elapsed_time": "1:14:55", "remaining_time": "1:22:24"}
|
| 151 |
+
{"current_steps": 151, "total_steps": 315, "loss": 0.7981, "lr": 6.289626849272062e-06, "epoch": 2.3968253968253967, "percentage": 47.94, "elapsed_time": "1:15:23", "remaining_time": "1:21:53"}
|
| 152 |
+
{"current_steps": 152, "total_steps": 315, "loss": 0.751, "lr": 6.2359213448636104e-06, "epoch": 2.4126984126984126, "percentage": 48.25, "elapsed_time": "1:15:52", "remaining_time": "1:21:21"}
|
| 153 |
+
{"current_steps": 153, "total_steps": 315, "loss": 0.7654, "lr": 6.182063535775634e-06, "epoch": 2.4285714285714284, "percentage": 48.57, "elapsed_time": "1:16:21", "remaining_time": "1:20:50"}
|
| 154 |
+
{"current_steps": 154, "total_steps": 315, "loss": 0.7896, "lr": 6.1280600589971225e-06, "epoch": 2.4444444444444446, "percentage": 48.89, "elapsed_time": "1:16:50", "remaining_time": "1:20:19"}
|
| 155 |
+
{"current_steps": 155, "total_steps": 315, "loss": 0.8051, "lr": 6.073917569467934e-06, "epoch": 2.4603174603174605, "percentage": 49.21, "elapsed_time": "1:17:18", "remaining_time": "1:19:48"}
|
| 156 |
+
{"current_steps": 156, "total_steps": 315, "loss": 0.7038, "lr": 6.0196427392587085e-06, "epoch": 2.4761904761904763, "percentage": 49.52, "elapsed_time": "1:17:48", "remaining_time": "1:19:18"}
|
| 157 |
+
{"current_steps": 157, "total_steps": 315, "loss": 0.7422, "lr": 5.96524225674865e-06, "epoch": 2.492063492063492, "percentage": 49.84, "elapsed_time": "1:18:17", "remaining_time": "1:18:46"}
|
| 158 |
+
{"current_steps": 158, "total_steps": 315, "loss": 0.7053, "lr": 5.9107228258013085e-06, "epoch": 2.507936507936508, "percentage": 50.16, "elapsed_time": "1:18:45", "remaining_time": "1:18:15"}
|
| 159 |
+
{"current_steps": 159, "total_steps": 315, "loss": 0.6523, "lr": 5.856091164938451e-06, "epoch": 2.5238095238095237, "percentage": 50.48, "elapsed_time": "1:19:14", "remaining_time": "1:17:44"}
|
| 160 |
+
{"current_steps": 160, "total_steps": 315, "loss": 0.6895, "lr": 5.801354006512127e-06, "epoch": 2.5396825396825395, "percentage": 50.79, "elapsed_time": "1:19:43", "remaining_time": "1:17:13"}
|
| 161 |
+
{"current_steps": 161, "total_steps": 315, "loss": 0.6973, "lr": 5.746518095875033e-06, "epoch": 2.5555555555555554, "percentage": 51.11, "elapsed_time": "1:20:11", "remaining_time": "1:16:42"}
|
| 162 |
+
{"current_steps": 162, "total_steps": 315, "loss": 0.6285, "lr": 5.6915901905492586e-06, "epoch": 2.571428571428571, "percentage": 51.43, "elapsed_time": "1:20:40", "remaining_time": "1:16:11"}
|
| 163 |
+
{"current_steps": 163, "total_steps": 315, "loss": 0.5907, "lr": 5.6365770593935665e-06, "epoch": 2.5873015873015874, "percentage": 51.75, "elapsed_time": "1:21:00", "remaining_time": "1:15:32"}
|
| 164 |
+
{"current_steps": 164, "total_steps": 315, "loss": 0.7181, "lr": 5.581485481769231e-06, "epoch": 2.6031746031746033, "percentage": 52.06, "elapsed_time": "1:21:29", "remaining_time": "1:15:01"}
|
| 165 |
+
{"current_steps": 165, "total_steps": 315, "loss": 0.7978, "lr": 5.526322246704628e-06, "epoch": 2.619047619047619, "percentage": 52.38, "elapsed_time": "1:21:58", "remaining_time": "1:14:31"}
|
| 166 |
+
{"current_steps": 166, "total_steps": 315, "loss": 0.681, "lr": 5.471094152058592e-06, "epoch": 2.634920634920635, "percentage": 52.7, "elapsed_time": "1:22:26", "remaining_time": "1:14:00"}
|
| 167 |
+
{"current_steps": 167, "total_steps": 315, "loss": 0.7308, "lr": 5.415808003682717e-06, "epoch": 2.6507936507936507, "percentage": 53.02, "elapsed_time": "1:22:55", "remaining_time": "1:13:29"}
|
| 168 |
+
{"current_steps": 168, "total_steps": 315, "loss": 0.7136, "lr": 5.360470614582661e-06, "epoch": 2.6666666666666665, "percentage": 53.33, "elapsed_time": "1:23:24", "remaining_time": "1:12:59"}
|
| 169 |
+
{"current_steps": 169, "total_steps": 315, "loss": 0.7333, "lr": 5.305088804078559e-06, "epoch": 2.682539682539683, "percentage": 53.65, "elapsed_time": "1:23:53", "remaining_time": "1:12:28"}
|
| 170 |
+
{"current_steps": 170, "total_steps": 315, "loss": 0.6349, "lr": 5.249669396964665e-06, "epoch": 2.6984126984126986, "percentage": 53.97, "elapsed_time": "1:24:21", "remaining_time": "1:11:57"}
|
| 171 |
+
{"current_steps": 171, "total_steps": 315, "loss": 0.776, "lr": 5.1942192226683385e-06, "epoch": 2.7142857142857144, "percentage": 54.29, "elapsed_time": "1:24:50", "remaining_time": "1:11:26"}
|
| 172 |
+
{"current_steps": 172, "total_steps": 315, "loss": 0.5998, "lr": 5.138745114408427e-06, "epoch": 2.7301587301587302, "percentage": 54.6, "elapsed_time": "1:25:18", "remaining_time": "1:10:55"}
|
| 173 |
+
{"current_steps": 173, "total_steps": 315, "loss": 0.6676, "lr": 5.083253908353193e-06, "epoch": 2.746031746031746, "percentage": 54.92, "elapsed_time": "1:25:46", "remaining_time": "1:10:24"}
|
| 174 |
+
{"current_steps": 174, "total_steps": 315, "loss": 0.7831, "lr": 5.0277524427778986e-06, "epoch": 2.761904761904762, "percentage": 55.24, "elapsed_time": "1:26:16", "remaining_time": "1:09:54"}
|
| 175 |
+
{"current_steps": 175, "total_steps": 315, "loss": 0.7164, "lr": 4.972247557222102e-06, "epoch": 2.7777777777777777, "percentage": 55.56, "elapsed_time": "1:26:44", "remaining_time": "1:09:23"}
|
| 176 |
+
{"current_steps": 176, "total_steps": 315, "loss": 0.6805, "lr": 4.916746091646808e-06, "epoch": 2.7936507936507935, "percentage": 55.87, "elapsed_time": "1:27:13", "remaining_time": "1:08:53"}
|
| 177 |
+
{"current_steps": 177, "total_steps": 315, "loss": 0.724, "lr": 4.8612548855915755e-06, "epoch": 2.8095238095238093, "percentage": 56.19, "elapsed_time": "1:27:42", "remaining_time": "1:08:22"}
|
| 178 |
+
{"current_steps": 178, "total_steps": 315, "loss": 0.7446, "lr": 4.805780777331662e-06, "epoch": 2.825396825396825, "percentage": 56.51, "elapsed_time": "1:28:10", "remaining_time": "1:07:52"}
|
| 179 |
+
{"current_steps": 179, "total_steps": 315, "loss": 0.7124, "lr": 4.750330603035336e-06, "epoch": 2.8412698412698414, "percentage": 56.83, "elapsed_time": "1:28:39", "remaining_time": "1:07:21"}
|
| 180 |
+
{"current_steps": 180, "total_steps": 315, "loss": 0.7252, "lr": 4.694911195921443e-06, "epoch": 2.857142857142857, "percentage": 57.14, "elapsed_time": "1:29:08", "remaining_time": "1:06:51"}
|
| 181 |
+
{"current_steps": 181, "total_steps": 315, "loss": 0.6053, "lr": 4.6395293854173395e-06, "epoch": 2.873015873015873, "percentage": 57.46, "elapsed_time": "1:29:37", "remaining_time": "1:06:20"}
|
| 182 |
+
{"current_steps": 182, "total_steps": 315, "loss": 0.6828, "lr": 4.584191996317285e-06, "epoch": 2.888888888888889, "percentage": 57.78, "elapsed_time": "1:30:03", "remaining_time": "1:05:48"}
|
| 183 |
+
{"current_steps": 183, "total_steps": 315, "loss": 0.8414, "lr": 4.528905847941411e-06, "epoch": 2.9047619047619047, "percentage": 58.1, "elapsed_time": "1:30:33", "remaining_time": "1:05:19"}
|
| 184 |
+
{"current_steps": 184, "total_steps": 315, "loss": 0.6592, "lr": 4.473677753295375e-06, "epoch": 2.9206349206349205, "percentage": 58.41, "elapsed_time": "1:31:00", "remaining_time": "1:04:47"}
|
| 185 |
+
{"current_steps": 185, "total_steps": 315, "loss": 0.7122, "lr": 4.418514518230769e-06, "epoch": 2.9365079365079367, "percentage": 58.73, "elapsed_time": "1:31:29", "remaining_time": "1:04:17"}
|
| 186 |
+
{"current_steps": 186, "total_steps": 315, "loss": 0.7454, "lr": 4.363422940606435e-06, "epoch": 2.9523809523809526, "percentage": 59.05, "elapsed_time": "1:31:58", "remaining_time": "1:03:47"}
|
| 187 |
+
{"current_steps": 187, "total_steps": 315, "loss": 0.7621, "lr": 4.308409809450742e-06, "epoch": 2.9682539682539684, "percentage": 59.37, "elapsed_time": "1:32:27", "remaining_time": "1:03:17"}
|
| 188 |
+
{"current_steps": 188, "total_steps": 315, "loss": 0.7331, "lr": 4.253481904124968e-06, "epoch": 2.984126984126984, "percentage": 59.68, "elapsed_time": "1:32:55", "remaining_time": "1:02:46"}
|
| 189 |
+
{"current_steps": 189, "total_steps": 315, "loss": 0.604, "lr": 4.198645993487872e-06, "epoch": 3.0, "percentage": 60.0, "elapsed_time": "1:33:24", "remaining_time": "1:02:16"}
|
| 190 |
+
{"current_steps": 190, "total_steps": 315, "loss": 0.6844, "lr": 4.143908835061551e-06, "epoch": 3.015873015873016, "percentage": 60.32, "elapsed_time": "1:35:43", "remaining_time": "1:02:58"}
|
| 191 |
+
{"current_steps": 191, "total_steps": 315, "loss": 0.7, "lr": 4.089277174198694e-06, "epoch": 3.0317460317460316, "percentage": 60.63, "elapsed_time": "1:36:12", "remaining_time": "1:02:27"}
|
| 192 |
+
{"current_steps": 192, "total_steps": 315, "loss": 0.743, "lr": 4.0347577432513515e-06, "epoch": 3.0476190476190474, "percentage": 60.95, "elapsed_time": "1:36:41", "remaining_time": "1:01:56"}
|
| 193 |
+
{"current_steps": 193, "total_steps": 315, "loss": 0.6811, "lr": 3.980357260741293e-06, "epoch": 3.0634920634920633, "percentage": 61.27, "elapsed_time": "1:37:10", "remaining_time": "1:01:25"}
|
| 194 |
+
{"current_steps": 194, "total_steps": 315, "loss": 0.6461, "lr": 3.926082430532067e-06, "epoch": 3.0793650793650795, "percentage": 61.59, "elapsed_time": "1:37:39", "remaining_time": "1:00:54"}
|
| 195 |
+
{"current_steps": 195, "total_steps": 315, "loss": 0.6076, "lr": 3.87193994100288e-06, "epoch": 3.0952380952380953, "percentage": 61.9, "elapsed_time": "1:38:07", "remaining_time": "1:00:23"}
|
| 196 |
+
{"current_steps": 196, "total_steps": 315, "loss": 0.6729, "lr": 3.817936464224367e-06, "epoch": 3.111111111111111, "percentage": 62.22, "elapsed_time": "1:38:37", "remaining_time": "0:59:52"}
|
| 197 |
+
{"current_steps": 197, "total_steps": 315, "loss": 0.7462, "lr": 3.764078655136391e-06, "epoch": 3.126984126984127, "percentage": 62.54, "elapsed_time": "1:39:05", "remaining_time": "0:59:21"}
|
| 198 |
+
{"current_steps": 198, "total_steps": 315, "loss": 0.6994, "lr": 3.7103731507279383e-06, "epoch": 3.142857142857143, "percentage": 62.86, "elapsed_time": "1:39:34", "remaining_time": "0:58:50"}
|
| 199 |
+
{"current_steps": 199, "total_steps": 315, "loss": 0.6698, "lr": 3.656826569219233e-06, "epoch": 3.1587301587301586, "percentage": 63.17, "elapsed_time": "1:40:02", "remaining_time": "0:58:19"}
|
| 200 |
+
{"current_steps": 200, "total_steps": 315, "loss": 0.641, "lr": 3.603445509246154e-06, "epoch": 3.1746031746031744, "percentage": 63.49, "elapsed_time": "1:40:31", "remaining_time": "0:57:47"}
|
| 201 |
+
{"current_steps": 201, "total_steps": 315, "loss": 0.7151, "lr": 3.55023654904709e-06, "epoch": 3.1904761904761907, "percentage": 63.81, "elapsed_time": "1:41:00", "remaining_time": "0:57:17"}
|
| 202 |
+
{"current_steps": 202, "total_steps": 315, "loss": 0.5512, "lr": 3.49720624565228e-06, "epoch": 3.2063492063492065, "percentage": 64.13, "elapsed_time": "1:41:29", "remaining_time": "0:56:46"}
|
| 203 |
+
{"current_steps": 203, "total_steps": 315, "loss": 0.6851, "lr": 3.44436113407578e-06, "epoch": 3.2222222222222223, "percentage": 64.44, "elapsed_time": "1:41:58", "remaining_time": "0:56:15"}
|
| 204 |
+
{"current_steps": 204, "total_steps": 315, "loss": 0.7484, "lr": 3.3917077265101505e-06, "epoch": 3.238095238095238, "percentage": 64.76, "elapsed_time": "1:42:27", "remaining_time": "0:55:44"}
|
| 205 |
+
{"current_steps": 205, "total_steps": 315, "loss": 0.5734, "lr": 3.3392525115239353e-06, "epoch": 3.253968253968254, "percentage": 65.08, "elapsed_time": "1:42:53", "remaining_time": "0:55:12"}
|
| 206 |
+
{"current_steps": 206, "total_steps": 315, "loss": 0.7111, "lr": 3.2870019532620744e-06, "epoch": 3.2698412698412698, "percentage": 65.4, "elapsed_time": "1:43:22", "remaining_time": "0:54:41"}
|
| 207 |
+
{"current_steps": 207, "total_steps": 315, "loss": 0.6547, "lr": 3.2349624906493164e-06, "epoch": 3.2857142857142856, "percentage": 65.71, "elapsed_time": "1:43:50", "remaining_time": "0:54:10"}
|
| 208 |
+
{"current_steps": 208, "total_steps": 315, "loss": 0.5268, "lr": 3.1831405365967315e-06, "epoch": 3.3015873015873014, "percentage": 66.03, "elapsed_time": "1:44:19", "remaining_time": "0:53:39"}
|
| 209 |
+
{"current_steps": 209, "total_steps": 315, "loss": 0.7138, "lr": 3.1315424772114404e-06, "epoch": 3.317460317460317, "percentage": 66.35, "elapsed_time": "1:44:49", "remaining_time": "0:53:09"}
|
| 210 |
+
{"current_steps": 210, "total_steps": 315, "loss": 0.5523, "lr": 3.0801746710096497e-06, "epoch": 3.3333333333333335, "percentage": 66.67, "elapsed_time": "1:45:13", "remaining_time": "0:52:36"}
|
| 211 |
+
{"current_steps": 211, "total_steps": 315, "loss": 0.586, "lr": 3.0290434481330746e-06, "epoch": 3.3492063492063493, "percentage": 66.98, "elapsed_time": "1:45:42", "remaining_time": "0:52:05"}
|
| 212 |
+
{"current_steps": 212, "total_steps": 315, "loss": 0.6185, "lr": 2.978155109568864e-06, "epoch": 3.365079365079365, "percentage": 67.3, "elapsed_time": "1:46:10", "remaining_time": "0:51:35"}
|
| 213 |
+
{"current_steps": 213, "total_steps": 315, "loss": 0.6636, "lr": 2.927515926373129e-06, "epoch": 3.380952380952381, "percentage": 67.62, "elapsed_time": "1:46:39", "remaining_time": "0:51:04"}
|
| 214 |
+
{"current_steps": 214, "total_steps": 315, "loss": 0.6636, "lr": 2.8771321388981334e-06, "epoch": 3.3968253968253967, "percentage": 67.94, "elapsed_time": "1:47:08", "remaining_time": "0:50:33"}
|
| 215 |
+
{"current_steps": 215, "total_steps": 315, "loss": 0.7087, "lr": 2.8270099560232992e-06, "epoch": 3.4126984126984126, "percentage": 68.25, "elapsed_time": "1:47:36", "remaining_time": "0:50:03"}
|
| 216 |
+
{"current_steps": 216, "total_steps": 315, "loss": 0.7505, "lr": 2.77715555439007e-06, "epoch": 3.4285714285714284, "percentage": 68.57, "elapsed_time": "1:48:05", "remaining_time": "0:49:32"}
|
| 217 |
+
{"current_steps": 217, "total_steps": 315, "loss": 0.6028, "lr": 2.7275750776407568e-06, "epoch": 3.4444444444444446, "percentage": 68.89, "elapsed_time": "1:48:33", "remaining_time": "0:49:01"}
|
| 218 |
+
{"current_steps": 218, "total_steps": 315, "loss": 0.5843, "lr": 2.6782746356614364e-06, "epoch": 3.4603174603174605, "percentage": 69.21, "elapsed_time": "1:49:01", "remaining_time": "0:48:30"}
|
| 219 |
+
{"current_steps": 219, "total_steps": 315, "loss": 0.7137, "lr": 2.6292603038290306e-06, "epoch": 3.4761904761904763, "percentage": 69.52, "elapsed_time": "1:49:30", "remaining_time": "0:48:00"}
|
| 220 |
+
{"current_steps": 220, "total_steps": 315, "loss": 0.6527, "lr": 2.580538122262627e-06, "epoch": 3.492063492063492, "percentage": 69.84, "elapsed_time": "1:49:59", "remaining_time": "0:47:29"}
|
| 221 |
+
{"current_steps": 221, "total_steps": 315, "loss": 0.5754, "lr": 2.532114095079137e-06, "epoch": 3.507936507936508, "percentage": 70.16, "elapsed_time": "1:50:27", "remaining_time": "0:46:58"}
|
| 222 |
+
{"current_steps": 222, "total_steps": 315, "loss": 0.6267, "lr": 2.4839941896534027e-06, "epoch": 3.5238095238095237, "percentage": 70.48, "elapsed_time": "1:50:56", "remaining_time": "0:46:28"}
|
| 223 |
+
{"current_steps": 223, "total_steps": 315, "loss": 0.5873, "lr": 2.4361843358828287e-06, "epoch": 3.5396825396825395, "percentage": 70.79, "elapsed_time": "1:51:24", "remaining_time": "0:45:57"}
|
| 224 |
+
{"current_steps": 224, "total_steps": 315, "loss": 0.7168, "lr": 2.388690425456629e-06, "epoch": 3.5555555555555554, "percentage": 71.11, "elapsed_time": "1:51:54", "remaining_time": "0:45:27"}
|
| 225 |
+
{"current_steps": 225, "total_steps": 315, "loss": 0.5671, "lr": 2.341518311129781e-06, "epoch": 3.571428571428571, "percentage": 71.43, "elapsed_time": "1:52:21", "remaining_time": "0:44:56"}
|
| 226 |
+
{"current_steps": 226, "total_steps": 315, "loss": 0.708, "lr": 2.2946738060017947e-06, "epoch": 3.5873015873015874, "percentage": 71.75, "elapsed_time": "1:52:49", "remaining_time": "0:44:25"}
|
| 227 |
+
{"current_steps": 227, "total_steps": 315, "loss": 0.653, "lr": 2.24816268280033e-06, "epoch": 3.6031746031746033, "percentage": 72.06, "elapsed_time": "1:53:18", "remaining_time": "0:43:55"}
|
| 228 |
+
{"current_steps": 228, "total_steps": 315, "loss": 0.6494, "lr": 2.2019906731698337e-06, "epoch": 3.619047619047619, "percentage": 72.38, "elapsed_time": "1:53:46", "remaining_time": "0:43:24"}
|
| 229 |
+
{"current_steps": 229, "total_steps": 315, "loss": 0.6293, "lr": 2.156163466965218e-06, "epoch": 3.634920634920635, "percentage": 72.7, "elapsed_time": "1:54:15", "remaining_time": "0:42:54"}
|
| 230 |
+
{"current_steps": 230, "total_steps": 315, "loss": 0.7077, "lr": 2.110686711550678e-06, "epoch": 3.6507936507936507, "percentage": 73.02, "elapsed_time": "1:54:40", "remaining_time": "0:42:22"}
|
| 231 |
+
{"current_steps": 231, "total_steps": 315, "loss": 0.6339, "lr": 2.0655660111037685e-06, "epoch": 3.6666666666666665, "percentage": 73.33, "elapsed_time": "1:55:08", "remaining_time": "0:41:52"}
|
| 232 |
+
{"current_steps": 232, "total_steps": 315, "loss": 0.6577, "lr": 2.0208069259247866e-06, "epoch": 3.682539682539683, "percentage": 73.65, "elapsed_time": "1:55:38", "remaining_time": "0:41:22"}
|
| 233 |
+
{"current_steps": 233, "total_steps": 315, "loss": 0.6029, "lr": 1.976414971751568e-06, "epoch": 3.6984126984126986, "percentage": 73.97, "elapsed_time": "1:56:07", "remaining_time": "0:40:51"}
|
| 234 |
+
{"current_steps": 234, "total_steps": 315, "loss": 0.6777, "lr": 1.932395619079771e-06, "epoch": 3.7142857142857144, "percentage": 74.29, "elapsed_time": "1:56:35", "remaining_time": "0:40:21"}
|
| 235 |
+
{"current_steps": 235, "total_steps": 315, "loss": 0.6996, "lr": 1.8887542924887486e-06, "epoch": 3.7301587301587302, "percentage": 74.6, "elapsed_time": "1:57:05", "remaining_time": "0:39:51"}
|
| 236 |
+
{"current_steps": 236, "total_steps": 315, "loss": 0.6727, "lr": 1.8454963699730471e-06, "epoch": 3.746031746031746, "percentage": 74.92, "elapsed_time": "1:57:34", "remaining_time": "0:39:21"}
|
| 237 |
+
{"current_steps": 237, "total_steps": 315, "loss": 0.5927, "lr": 1.802627182279687e-06, "epoch": 3.761904761904762, "percentage": 75.24, "elapsed_time": "1:58:01", "remaining_time": "0:38:50"}
|
| 238 |
+
{"current_steps": 238, "total_steps": 315, "loss": 0.5039, "lr": 1.760152012251241e-06, "epoch": 3.7777777777777777, "percentage": 75.56, "elapsed_time": "1:58:30", "remaining_time": "0:38:20"}
|
| 239 |
+
{"current_steps": 239, "total_steps": 315, "loss": 0.6682, "lr": 1.7180760941748132e-06, "epoch": 3.7936507936507935, "percentage": 75.87, "elapsed_time": "1:58:59", "remaining_time": "0:37:50"}
|
| 240 |
+
{"current_steps": 240, "total_steps": 315, "loss": 0.7302, "lr": 1.6764046131370142e-06, "epoch": 3.8095238095238093, "percentage": 76.19, "elapsed_time": "1:59:27", "remaining_time": "0:37:19"}
|
| 241 |
+
{"current_steps": 241, "total_steps": 315, "loss": 0.6953, "lr": 1.6351427043849955e-06, "epoch": 3.825396825396825, "percentage": 76.51, "elapsed_time": "1:59:56", "remaining_time": "0:36:49"}
|
| 242 |
+
{"current_steps": 242, "total_steps": 315, "loss": 0.654, "lr": 1.5942954526936217e-06, "epoch": 3.8412698412698414, "percentage": 76.83, "elapsed_time": "2:00:24", "remaining_time": "0:36:19"}
|
| 243 |
+
{"current_steps": 243, "total_steps": 315, "loss": 0.6618, "lr": 1.5538678917388638e-06, "epoch": 3.857142857142857, "percentage": 77.14, "elapsed_time": "2:00:53", "remaining_time": "0:35:49"}
|
| 244 |
+
{"current_steps": 244, "total_steps": 315, "loss": 0.5709, "lr": 1.5138650034775004e-06, "epoch": 3.873015873015873, "percentage": 77.46, "elapsed_time": "2:01:22", "remaining_time": "0:35:19"}
|
| 245 |
+
{"current_steps": 245, "total_steps": 315, "loss": 0.7154, "lr": 1.4742917175331644e-06, "epoch": 3.888888888888889, "percentage": 77.78, "elapsed_time": "2:01:51", "remaining_time": "0:34:48"}
|
| 246 |
+
{"current_steps": 246, "total_steps": 315, "loss": 0.7655, "lr": 1.4351529105888735e-06, "epoch": 3.9047619047619047, "percentage": 78.1, "elapsed_time": "2:02:20", "remaining_time": "0:34:18"}
|
| 247 |
+
{"current_steps": 247, "total_steps": 315, "loss": 0.5952, "lr": 1.3964534057860652e-06, "epoch": 3.9206349206349205, "percentage": 78.41, "elapsed_time": "2:02:48", "remaining_time": "0:33:48"}
|
| 248 |
+
{"current_steps": 248, "total_steps": 315, "loss": 0.6536, "lr": 1.3581979721302286e-06, "epoch": 3.9365079365079367, "percentage": 78.73, "elapsed_time": "2:03:17", "remaining_time": "0:33:18"}
|
| 249 |
+
{"current_steps": 249, "total_steps": 315, "loss": 0.6675, "lr": 1.3203913239032074e-06, "epoch": 3.9523809523809526, "percentage": 79.05, "elapsed_time": "2:03:46", "remaining_time": "0:32:48"}
|
| 250 |
+
{"current_steps": 250, "total_steps": 315, "loss": 0.6196, "lr": 1.283038120082268e-06, "epoch": 3.9682539682539684, "percentage": 79.37, "elapsed_time": "2:04:15", "remaining_time": "0:32:18"}
|
| 251 |
+
{"current_steps": 251, "total_steps": 315, "loss": 0.6194, "lr": 1.2461429637659466e-06, "epoch": 3.984126984126984, "percentage": 79.68, "elapsed_time": "2:04:43", "remaining_time": "0:31:48"}
|
| 252 |
+
{"current_steps": 252, "total_steps": 315, "loss": 0.6338, "lr": 1.2097104016068146e-06, "epoch": 4.0, "percentage": 80.0, "elapsed_time": "2:05:14", "remaining_time": "0:31:18"}
|
| 253 |
+
{"current_steps": 253, "total_steps": 315, "loss": 0.6351, "lr": 1.1737449232511799e-06, "epoch": 4.015873015873016, "percentage": 80.32, "elapsed_time": "2:07:42", "remaining_time": "0:31:17"}
|
| 254 |
+
{"current_steps": 254, "total_steps": 315, "loss": 0.6771, "lr": 1.1382509607858233e-06, "epoch": 4.031746031746032, "percentage": 80.63, "elapsed_time": "2:08:11", "remaining_time": "0:30:47"}
|
| 255 |
+
{"current_steps": 255, "total_steps": 315, "loss": 0.6626, "lr": 1.1032328881918237e-06, "epoch": 4.0476190476190474, "percentage": 80.95, "elapsed_time": "2:08:40", "remaining_time": "0:30:16"}
|
| 256 |
+
{"current_steps": 256, "total_steps": 315, "loss": 0.6944, "lr": 1.0686950208055486e-06, "epoch": 4.063492063492063, "percentage": 81.27, "elapsed_time": "2:09:08", "remaining_time": "0:29:45"}
|
| 257 |
+
{"current_steps": 257, "total_steps": 315, "loss": 0.6248, "lr": 1.034641614786862e-06, "epoch": 4.079365079365079, "percentage": 81.59, "elapsed_time": "2:09:37", "remaining_time": "0:29:15"}
|
| 258 |
+
{"current_steps": 258, "total_steps": 315, "loss": 0.6044, "lr": 1.0010768665946309e-06, "epoch": 4.095238095238095, "percentage": 81.9, "elapsed_time": "2:10:05", "remaining_time": "0:28:44"}
|
| 259 |
+
{"current_steps": 259, "total_steps": 315, "loss": 0.6343, "lr": 9.680049124695973e-07, "epoch": 4.111111111111111, "percentage": 82.22, "elapsed_time": "2:10:35", "remaining_time": "0:28:14"}
|
| 260 |
+
{"current_steps": 260, "total_steps": 315, "loss": 0.6539, "lr": 9.35429827924652e-07, "epoch": 4.1269841269841265, "percentage": 82.54, "elapsed_time": "2:11:04", "remaining_time": "0:27:43"}
|
| 261 |
+
{"current_steps": 261, "total_steps": 315, "loss": 0.5768, "lr": 9.033556272426075e-07, "epoch": 4.142857142857143, "percentage": 82.86, "elapsed_time": "2:11:34", "remaining_time": "0:27:13"}
|
| 262 |
+
{"current_steps": 262, "total_steps": 315, "loss": 0.66, "lr": 8.717862629815099e-07, "epoch": 4.158730158730159, "percentage": 83.17, "elapsed_time": "2:12:03", "remaining_time": "0:26:42"}
|
| 263 |
+
{"current_steps": 263, "total_steps": 315, "loss": 0.5534, "lr": 8.407256254875573e-07, "epoch": 4.174603174603175, "percentage": 83.49, "elapsed_time": "2:12:31", "remaining_time": "0:26:12"}
|
| 264 |
+
{"current_steps": 264, "total_steps": 315, "loss": 0.7396, "lr": 8.101775424156888e-07, "epoch": 4.190476190476191, "percentage": 83.81, "elapsed_time": "2:13:00", "remaining_time": "0:25:41"}
|
| 265 |
+
{"current_steps": 265, "total_steps": 315, "loss": 0.574, "lr": 7.801457782578947e-07, "epoch": 4.2063492063492065, "percentage": 84.13, "elapsed_time": "2:13:29", "remaining_time": "0:25:11"}
|
| 266 |
+
{"current_steps": 266, "total_steps": 315, "loss": 0.6995, "lr": 7.506340338793111e-07, "epoch": 4.222222222222222, "percentage": 84.44, "elapsed_time": "2:13:59", "remaining_time": "0:24:40"}
|
| 267 |
+
{"current_steps": 267, "total_steps": 315, "loss": 0.5681, "lr": 7.216459460621528e-07, "epoch": 4.238095238095238, "percentage": 84.76, "elapsed_time": "2:14:28", "remaining_time": "0:24:10"}
|
| 268 |
+
{"current_steps": 268, "total_steps": 315, "loss": 0.6041, "lr": 6.931850870575563e-07, "epoch": 4.253968253968254, "percentage": 85.08, "elapsed_time": "2:14:56", "remaining_time": "0:23:39"}
|
| 269 |
+
{"current_steps": 269, "total_steps": 315, "loss": 0.652, "lr": 6.652549641453543e-07, "epoch": 4.26984126984127, "percentage": 85.4, "elapsed_time": "2:15:25", "remaining_time": "0:23:09"}
|
| 270 |
+
{"current_steps": 270, "total_steps": 315, "loss": 0.5264, "lr": 6.378590192018752e-07, "epoch": 4.285714285714286, "percentage": 85.71, "elapsed_time": "2:15:52", "remaining_time": "0:22:38"}
|
| 271 |
+
{"current_steps": 271, "total_steps": 315, "loss": 0.6315, "lr": 6.110006282757897e-07, "epoch": 4.301587301587301, "percentage": 86.03, "elapsed_time": "2:16:21", "remaining_time": "0:22:08"}
|
| 272 |
+
{"current_steps": 272, "total_steps": 315, "loss": 0.5652, "lr": 5.846831011720789e-07, "epoch": 4.317460317460317, "percentage": 86.35, "elapsed_time": "2:16:48", "remaining_time": "0:21:37"}
|
| 273 |
+
{"current_steps": 273, "total_steps": 315, "loss": 0.5355, "lr": 5.589096810441574e-07, "epoch": 4.333333333333333, "percentage": 86.67, "elapsed_time": "2:17:15", "remaining_time": "0:21:07"}
|
| 274 |
+
{"current_steps": 274, "total_steps": 315, "loss": 0.6575, "lr": 5.3368354399422e-07, "epoch": 4.349206349206349, "percentage": 86.98, "elapsed_time": "2:17:44", "remaining_time": "0:20:36"}
|
| 275 |
+
{"current_steps": 275, "total_steps": 315, "loss": 0.5854, "lr": 5.090077986818365e-07, "epoch": 4.365079365079365, "percentage": 87.3, "elapsed_time": "2:18:12", "remaining_time": "0:20:06"}
|
| 276 |
+
{"current_steps": 276, "total_steps": 315, "loss": 0.6625, "lr": 4.848854859408731e-07, "epoch": 4.380952380952381, "percentage": 87.62, "elapsed_time": "2:18:41", "remaining_time": "0:19:35"}
|
| 277 |
+
{"current_steps": 277, "total_steps": 315, "loss": 0.596, "lr": 4.613195784047653e-07, "epoch": 4.396825396825397, "percentage": 87.94, "elapsed_time": "2:19:10", "remaining_time": "0:19:05"}
|
| 278 |
+
{"current_steps": 278, "total_steps": 315, "loss": 0.6383, "lr": 4.3831298014019144e-07, "epoch": 4.412698412698413, "percentage": 88.25, "elapsed_time": "2:19:39", "remaining_time": "0:18:35"}
|
| 279 |
+
{"current_steps": 279, "total_steps": 315, "loss": 0.5529, "lr": 4.1586852628920095e-07, "epoch": 4.428571428571429, "percentage": 88.57, "elapsed_time": "2:20:07", "remaining_time": "0:18:04"}
|
| 280 |
+
{"current_steps": 280, "total_steps": 315, "loss": 0.4942, "lr": 3.939889827198362e-07, "epoch": 4.444444444444445, "percentage": 88.89, "elapsed_time": "2:20:35", "remaining_time": "0:17:34"}
|
| 281 |
+
{"current_steps": 281, "total_steps": 315, "loss": 0.5257, "lr": 3.7267704568529015e-07, "epoch": 4.4603174603174605, "percentage": 89.21, "elapsed_time": "2:21:03", "remaining_time": "0:17:04"}
|
| 282 |
+
{"current_steps": 282, "total_steps": 315, "loss": 0.6823, "lr": 3.519353414916404e-07, "epoch": 4.476190476190476, "percentage": 89.52, "elapsed_time": "2:21:32", "remaining_time": "0:16:33"}
|
| 283 |
+
{"current_steps": 283, "total_steps": 315, "loss": 0.6414, "lr": 3.3176642617420817e-07, "epoch": 4.492063492063492, "percentage": 89.84, "elapsed_time": "2:22:01", "remaining_time": "0:16:03"}
|
| 284 |
+
{"current_steps": 284, "total_steps": 315, "loss": 0.6771, "lr": 3.1217278518256844e-07, "epoch": 4.507936507936508, "percentage": 90.16, "elapsed_time": "2:22:30", "remaining_time": "0:15:33"}
|
| 285 |
+
{"current_steps": 285, "total_steps": 315, "loss": 0.7079, "lr": 2.93156833074269e-07, "epoch": 4.523809523809524, "percentage": 90.48, "elapsed_time": "2:22:58", "remaining_time": "0:15:03"}
|
| 286 |
+
{"current_steps": 286, "total_steps": 315, "loss": 0.5173, "lr": 2.7472091321728067e-07, "epoch": 4.5396825396825395, "percentage": 90.79, "elapsed_time": "2:23:27", "remaining_time": "0:14:32"}
|
| 287 |
+
{"current_steps": 287, "total_steps": 315, "loss": 0.5528, "lr": 2.568672975012154e-07, "epoch": 4.555555555555555, "percentage": 91.11, "elapsed_time": "2:23:56", "remaining_time": "0:14:02"}
|
| 288 |
+
{"current_steps": 288, "total_steps": 315, "loss": 0.5663, "lr": 2.3959818605736095e-07, "epoch": 4.571428571428571, "percentage": 91.43, "elapsed_time": "2:24:24", "remaining_time": "0:13:32"}
|
| 289 |
+
{"current_steps": 289, "total_steps": 315, "loss": 0.6321, "lr": 2.229157069875537e-07, "epoch": 4.587301587301587, "percentage": 91.75, "elapsed_time": "2:24:53", "remaining_time": "0:13:02"}
|
| 290 |
+
{"current_steps": 290, "total_steps": 315, "loss": 0.4887, "lr": 2.068219161019297e-07, "epoch": 4.603174603174603, "percentage": 92.06, "elapsed_time": "2:25:18", "remaining_time": "0:12:31"}
|
| 291 |
+
{"current_steps": 291, "total_steps": 315, "loss": 0.6286, "lr": 1.9131879666558385e-07, "epoch": 4.619047619047619, "percentage": 92.38, "elapsed_time": "2:25:47", "remaining_time": "0:12:01"}
|
| 292 |
+
{"current_steps": 292, "total_steps": 315, "loss": 0.5375, "lr": 1.7640825915416994e-07, "epoch": 4.634920634920634, "percentage": 92.7, "elapsed_time": "2:26:15", "remaining_time": "0:11:31"}
|
| 293 |
+
{"current_steps": 293, "total_steps": 315, "loss": 0.5712, "lr": 1.6209214101846394e-07, "epoch": 4.650793650793651, "percentage": 93.02, "elapsed_time": "2:26:42", "remaining_time": "0:11:00"}
|
| 294 |
+
{"current_steps": 294, "total_steps": 315, "loss": 0.6848, "lr": 1.4837220645793905e-07, "epoch": 4.666666666666667, "percentage": 93.33, "elapsed_time": "2:27:11", "remaining_time": "0:10:30"}
|
| 295 |
+
{"current_steps": 295, "total_steps": 315, "loss": 0.671, "lr": 1.3525014620335786e-07, "epoch": 4.682539682539683, "percentage": 93.65, "elapsed_time": "2:27:40", "remaining_time": "0:10:00"}
|
| 296 |
+
{"current_steps": 296, "total_steps": 315, "loss": 0.7189, "lr": 1.2272757730841744e-07, "epoch": 4.698412698412699, "percentage": 93.97, "elapsed_time": "2:28:09", "remaining_time": "0:09:30"}
|
| 297 |
+
{"current_steps": 297, "total_steps": 315, "loss": 0.5655, "lr": 1.1080604295048203e-07, "epoch": 4.714285714285714, "percentage": 94.29, "elapsed_time": "2:28:37", "remaining_time": "0:09:00"}
|
| 298 |
+
{"current_steps": 298, "total_steps": 315, "loss": 0.678, "lr": 9.948701224041124e-08, "epoch": 4.73015873015873, "percentage": 94.6, "elapsed_time": "2:29:06", "remaining_time": "0:08:30"}
|
| 299 |
+
{"current_steps": 299, "total_steps": 315, "loss": 0.677, "lr": 8.877188004152104e-08, "epoch": 4.746031746031746, "percentage": 94.92, "elapsed_time": "2:29:34", "remaining_time": "0:08:00"}
|
| 300 |
+
{"current_steps": 300, "total_steps": 315, "loss": 0.5844, "lr": 7.866196679768956e-08, "epoch": 4.761904761904762, "percentage": 95.24, "elapsed_time": "2:30:03", "remaining_time": "0:07:30"}
|
| 301 |
+
{"current_steps": 301, "total_steps": 315, "loss": 0.5936, "lr": 6.91585183706428e-08, "epoch": 4.777777777777778, "percentage": 95.56, "elapsed_time": "2:30:32", "remaining_time": "0:07:00"}
|
| 302 |
+
{"current_steps": 302, "total_steps": 315, "loss": 0.6322, "lr": 6.02627058864158e-08, "epoch": 4.7936507936507935, "percentage": 95.87, "elapsed_time": "2:31:01", "remaining_time": "0:06:30"}
|
| 303 |
+
{"current_steps": 303, "total_steps": 315, "loss": 0.6218, "lr": 5.19756255910403e-08, "epoch": 4.809523809523809, "percentage": 96.19, "elapsed_time": "2:31:29", "remaining_time": "0:05:59"}
|
| 304 |
+
{"current_steps": 304, "total_steps": 315, "loss": 0.6154, "lr": 4.429829871545055e-08, "epoch": 4.825396825396825, "percentage": 96.51, "elapsed_time": "2:31:55", "remaining_time": "0:05:29"}
|
| 305 |
+
{"current_steps": 305, "total_steps": 315, "loss": 0.5363, "lr": 3.7231671349634015e-08, "epoch": 4.841269841269841, "percentage": 96.83, "elapsed_time": "2:32:19", "remaining_time": "0:04:59"}
|
| 306 |
+
{"current_steps": 306, "total_steps": 315, "loss": 0.5705, "lr": 3.077661432604184e-08, "epoch": 4.857142857142857, "percentage": 97.14, "elapsed_time": "2:32:47", "remaining_time": "0:04:29"}
|
| 307 |
+
{"current_steps": 307, "total_steps": 315, "loss": 0.6748, "lr": 2.4933923112279712e-08, "epoch": 4.8730158730158735, "percentage": 97.46, "elapsed_time": "2:33:16", "remaining_time": "0:03:59"}
|
| 308 |
+
{"current_steps": 308, "total_steps": 315, "loss": 0.7078, "lr": 1.9704317713076236e-08, "epoch": 4.888888888888889, "percentage": 97.78, "elapsed_time": "2:33:45", "remaining_time": "0:03:29"}
|
| 309 |
+
{"current_steps": 309, "total_steps": 315, "loss": 0.6406, "lr": 1.508844258155728e-08, "epoch": 4.904761904761905, "percentage": 98.1, "elapsed_time": "2:34:13", "remaining_time": "0:02:59"}
|
| 310 |
+
{"current_steps": 310, "total_steps": 315, "loss": 0.6818, "lr": 1.1086866539830044e-08, "epoch": 4.920634920634921, "percentage": 98.41, "elapsed_time": "2:34:42", "remaining_time": "0:02:29"}
|
| 311 |
+
{"current_steps": 311, "total_steps": 315, "loss": 0.6386, "lr": 7.700082708883006e-09, "epoch": 4.936507936507937, "percentage": 98.73, "elapsed_time": "2:35:11", "remaining_time": "0:01:59"}
|
| 312 |
+
{"current_steps": 312, "total_steps": 315, "loss": 0.6198, "lr": 4.928508447821223e-09, "epoch": 4.9523809523809526, "percentage": 99.05, "elapsed_time": "2:35:40", "remaining_time": "0:01:29"}
|
| 313 |
+
{"current_steps": 313, "total_steps": 315, "loss": 0.5728, "lr": 2.7724853024324594e-09, "epoch": 4.968253968253968, "percentage": 99.37, "elapsed_time": "2:36:07", "remaining_time": "0:00:59"}
|
| 314 |
+
{"current_steps": 314, "total_steps": 315, "loss": 0.6566, "lr": 1.2322789630997422e-09, "epoch": 4.984126984126984, "percentage": 99.68, "elapsed_time": "2:36:37", "remaining_time": "0:00:29"}
|
| 315 |
+
{"current_steps": 315, "total_steps": 315, "loss": 0.5612, "lr": 3.080792320564463e-10, "epoch": 5.0, "percentage": 100.0, "elapsed_time": "2:37:08", "remaining_time": "0:00:00"}
|
| 316 |
+
{"current_steps": 315, "total_steps": 315, "epoch": 5.0, "percentage": 100.0, "elapsed_time": "2:38:58", "remaining_time": "0:00:00"}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2248 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 5.0,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 315,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.015873015873015872,
|
| 14 |
+
"grad_norm": 2.3559653063882835,
|
| 15 |
+
"learning_rate": 0.0,
|
| 16 |
+
"loss": 1.0469,
|
| 17 |
+
"step": 1
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.031746031746031744,
|
| 21 |
+
"grad_norm": 2.29659253609106,
|
| 22 |
+
"learning_rate": 3.125e-07,
|
| 23 |
+
"loss": 0.9929,
|
| 24 |
+
"step": 2
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.047619047619047616,
|
| 28 |
+
"grad_norm": 2.8899785663629123,
|
| 29 |
+
"learning_rate": 6.25e-07,
|
| 30 |
+
"loss": 1.0189,
|
| 31 |
+
"step": 3
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.06349206349206349,
|
| 35 |
+
"grad_norm": 2.2527917765154153,
|
| 36 |
+
"learning_rate": 9.375000000000001e-07,
|
| 37 |
+
"loss": 0.9098,
|
| 38 |
+
"step": 4
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.07936507936507936,
|
| 42 |
+
"grad_norm": 2.2029116064708907,
|
| 43 |
+
"learning_rate": 1.25e-06,
|
| 44 |
+
"loss": 1.0462,
|
| 45 |
+
"step": 5
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.09523809523809523,
|
| 49 |
+
"grad_norm": 2.122312207060731,
|
| 50 |
+
"learning_rate": 1.5625e-06,
|
| 51 |
+
"loss": 0.9986,
|
| 52 |
+
"step": 6
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.1111111111111111,
|
| 56 |
+
"grad_norm": 2.045608321522422,
|
| 57 |
+
"learning_rate": 1.8750000000000003e-06,
|
| 58 |
+
"loss": 0.9554,
|
| 59 |
+
"step": 7
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.12698412698412698,
|
| 63 |
+
"grad_norm": 1.8619152723479657,
|
| 64 |
+
"learning_rate": 2.1875000000000002e-06,
|
| 65 |
+
"loss": 0.9522,
|
| 66 |
+
"step": 8
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.14285714285714285,
|
| 70 |
+
"grad_norm": 1.8663709940802706,
|
| 71 |
+
"learning_rate": 2.5e-06,
|
| 72 |
+
"loss": 0.9994,
|
| 73 |
+
"step": 9
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.15873015873015872,
|
| 77 |
+
"grad_norm": 1.756651050051264,
|
| 78 |
+
"learning_rate": 2.8125e-06,
|
| 79 |
+
"loss": 0.9373,
|
| 80 |
+
"step": 10
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.1746031746031746,
|
| 84 |
+
"grad_norm": 1.6604489437599113,
|
| 85 |
+
"learning_rate": 3.125e-06,
|
| 86 |
+
"loss": 0.8839,
|
| 87 |
+
"step": 11
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.19047619047619047,
|
| 91 |
+
"grad_norm": 1.292095856227553,
|
| 92 |
+
"learning_rate": 3.4375e-06,
|
| 93 |
+
"loss": 0.9907,
|
| 94 |
+
"step": 12
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.20634920634920634,
|
| 98 |
+
"grad_norm": 1.2693344466908103,
|
| 99 |
+
"learning_rate": 3.7500000000000005e-06,
|
| 100 |
+
"loss": 0.9322,
|
| 101 |
+
"step": 13
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.2222222222222222,
|
| 105 |
+
"grad_norm": 1.6344663181288221,
|
| 106 |
+
"learning_rate": 4.0625000000000005e-06,
|
| 107 |
+
"loss": 1.0934,
|
| 108 |
+
"step": 14
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.23809523809523808,
|
| 112 |
+
"grad_norm": 1.1511973531225708,
|
| 113 |
+
"learning_rate": 4.3750000000000005e-06,
|
| 114 |
+
"loss": 0.9771,
|
| 115 |
+
"step": 15
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.25396825396825395,
|
| 119 |
+
"grad_norm": 1.1464903643465947,
|
| 120 |
+
"learning_rate": 4.6875000000000004e-06,
|
| 121 |
+
"loss": 1.0442,
|
| 122 |
+
"step": 16
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.2698412698412698,
|
| 126 |
+
"grad_norm": 1.1549964376534243,
|
| 127 |
+
"learning_rate": 5e-06,
|
| 128 |
+
"loss": 0.949,
|
| 129 |
+
"step": 17
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.2857142857142857,
|
| 133 |
+
"grad_norm": 1.2909225755543452,
|
| 134 |
+
"learning_rate": 5.3125e-06,
|
| 135 |
+
"loss": 1.0098,
|
| 136 |
+
"step": 18
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.30158730158730157,
|
| 140 |
+
"grad_norm": 1.0413894105842352,
|
| 141 |
+
"learning_rate": 5.625e-06,
|
| 142 |
+
"loss": 0.7468,
|
| 143 |
+
"step": 19
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.31746031746031744,
|
| 147 |
+
"grad_norm": 1.2863204827157997,
|
| 148 |
+
"learning_rate": 5.9375e-06,
|
| 149 |
+
"loss": 0.9232,
|
| 150 |
+
"step": 20
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.3333333333333333,
|
| 154 |
+
"grad_norm": 0.901486908411037,
|
| 155 |
+
"learning_rate": 6.25e-06,
|
| 156 |
+
"loss": 0.9885,
|
| 157 |
+
"step": 21
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.3492063492063492,
|
| 161 |
+
"grad_norm": 1.0229681930848715,
|
| 162 |
+
"learning_rate": 6.5625e-06,
|
| 163 |
+
"loss": 1.0508,
|
| 164 |
+
"step": 22
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.36507936507936506,
|
| 168 |
+
"grad_norm": 1.0772270927236638,
|
| 169 |
+
"learning_rate": 6.875e-06,
|
| 170 |
+
"loss": 0.8728,
|
| 171 |
+
"step": 23
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.38095238095238093,
|
| 175 |
+
"grad_norm": 0.9393952588410857,
|
| 176 |
+
"learning_rate": 7.1875e-06,
|
| 177 |
+
"loss": 0.8349,
|
| 178 |
+
"step": 24
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.3968253968253968,
|
| 182 |
+
"grad_norm": 1.0822345499912303,
|
| 183 |
+
"learning_rate": 7.500000000000001e-06,
|
| 184 |
+
"loss": 1.0283,
|
| 185 |
+
"step": 25
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.4126984126984127,
|
| 189 |
+
"grad_norm": 0.8124841375138875,
|
| 190 |
+
"learning_rate": 7.8125e-06,
|
| 191 |
+
"loss": 0.8612,
|
| 192 |
+
"step": 26
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.42857142857142855,
|
| 196 |
+
"grad_norm": 0.7868328056966778,
|
| 197 |
+
"learning_rate": 8.125000000000001e-06,
|
| 198 |
+
"loss": 0.7232,
|
| 199 |
+
"step": 27
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.4444444444444444,
|
| 203 |
+
"grad_norm": 0.8283002452965974,
|
| 204 |
+
"learning_rate": 8.4375e-06,
|
| 205 |
+
"loss": 0.9752,
|
| 206 |
+
"step": 28
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.4603174603174603,
|
| 210 |
+
"grad_norm": 0.8117454313345658,
|
| 211 |
+
"learning_rate": 8.750000000000001e-06,
|
| 212 |
+
"loss": 0.9053,
|
| 213 |
+
"step": 29
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.47619047619047616,
|
| 217 |
+
"grad_norm": 0.8594519052279771,
|
| 218 |
+
"learning_rate": 9.0625e-06,
|
| 219 |
+
"loss": 0.9479,
|
| 220 |
+
"step": 30
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.49206349206349204,
|
| 224 |
+
"grad_norm": 0.9958158956912483,
|
| 225 |
+
"learning_rate": 9.375000000000001e-06,
|
| 226 |
+
"loss": 0.9188,
|
| 227 |
+
"step": 31
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.5079365079365079,
|
| 231 |
+
"grad_norm": 0.8918575253813723,
|
| 232 |
+
"learning_rate": 9.6875e-06,
|
| 233 |
+
"loss": 0.8847,
|
| 234 |
+
"step": 32
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.5238095238095238,
|
| 238 |
+
"grad_norm": 0.8125329449215294,
|
| 239 |
+
"learning_rate": 1e-05,
|
| 240 |
+
"loss": 1.0379,
|
| 241 |
+
"step": 33
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.5396825396825397,
|
| 245 |
+
"grad_norm": 0.7108806956906407,
|
| 246 |
+
"learning_rate": 9.999691920767945e-06,
|
| 247 |
+
"loss": 0.8376,
|
| 248 |
+
"step": 34
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.5555555555555556,
|
| 252 |
+
"grad_norm": 0.641926463787557,
|
| 253 |
+
"learning_rate": 9.998767721036901e-06,
|
| 254 |
+
"loss": 0.8241,
|
| 255 |
+
"step": 35
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.5714285714285714,
|
| 259 |
+
"grad_norm": 0.7048636529194373,
|
| 260 |
+
"learning_rate": 9.997227514697568e-06,
|
| 261 |
+
"loss": 0.9693,
|
| 262 |
+
"step": 36
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.5873015873015873,
|
| 266 |
+
"grad_norm": 0.6041864409794199,
|
| 267 |
+
"learning_rate": 9.99507149155218e-06,
|
| 268 |
+
"loss": 0.9839,
|
| 269 |
+
"step": 37
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.6031746031746031,
|
| 273 |
+
"grad_norm": 0.6529381186048961,
|
| 274 |
+
"learning_rate": 9.992299917291118e-06,
|
| 275 |
+
"loss": 0.8479,
|
| 276 |
+
"step": 38
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.6190476190476191,
|
| 280 |
+
"grad_norm": 0.7457758141141355,
|
| 281 |
+
"learning_rate": 9.98891313346017e-06,
|
| 282 |
+
"loss": 0.9095,
|
| 283 |
+
"step": 39
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.6349206349206349,
|
| 287 |
+
"grad_norm": 0.6700791615416641,
|
| 288 |
+
"learning_rate": 9.984911557418444e-06,
|
| 289 |
+
"loss": 0.7685,
|
| 290 |
+
"step": 40
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.6507936507936508,
|
| 294 |
+
"grad_norm": 0.6202447937301818,
|
| 295 |
+
"learning_rate": 9.980295682286924e-06,
|
| 296 |
+
"loss": 0.8387,
|
| 297 |
+
"step": 41
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.6666666666666666,
|
| 301 |
+
"grad_norm": 0.6888680420644837,
|
| 302 |
+
"learning_rate": 9.97506607688772e-06,
|
| 303 |
+
"loss": 0.9107,
|
| 304 |
+
"step": 42
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.6825396825396826,
|
| 308 |
+
"grad_norm": 0.5229452850388104,
|
| 309 |
+
"learning_rate": 9.969223385673958e-06,
|
| 310 |
+
"loss": 0.8308,
|
| 311 |
+
"step": 43
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.6984126984126984,
|
| 315 |
+
"grad_norm": 0.5679326043532053,
|
| 316 |
+
"learning_rate": 9.962768328650367e-06,
|
| 317 |
+
"loss": 0.7516,
|
| 318 |
+
"step": 44
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.7142857142857143,
|
| 322 |
+
"grad_norm": 0.5234412349262514,
|
| 323 |
+
"learning_rate": 9.95570170128455e-06,
|
| 324 |
+
"loss": 0.8443,
|
| 325 |
+
"step": 45
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.7301587301587301,
|
| 329 |
+
"grad_norm": 0.5148736685750067,
|
| 330 |
+
"learning_rate": 9.94802437440896e-06,
|
| 331 |
+
"loss": 0.7959,
|
| 332 |
+
"step": 46
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.746031746031746,
|
| 336 |
+
"grad_norm": 0.6223703419413371,
|
| 337 |
+
"learning_rate": 9.939737294113585e-06,
|
| 338 |
+
"loss": 0.8964,
|
| 339 |
+
"step": 47
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.7619047619047619,
|
| 343 |
+
"grad_norm": 0.4712938980573866,
|
| 344 |
+
"learning_rate": 9.930841481629358e-06,
|
| 345 |
+
"loss": 0.8884,
|
| 346 |
+
"step": 48
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.7777777777777778,
|
| 350 |
+
"grad_norm": 0.6385581101993485,
|
| 351 |
+
"learning_rate": 9.92133803320231e-06,
|
| 352 |
+
"loss": 0.7817,
|
| 353 |
+
"step": 49
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.7936507936507936,
|
| 357 |
+
"grad_norm": 0.47528095545287,
|
| 358 |
+
"learning_rate": 9.91122811995848e-06,
|
| 359 |
+
"loss": 0.819,
|
| 360 |
+
"step": 50
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.8095238095238095,
|
| 364 |
+
"grad_norm": 0.5522186664203698,
|
| 365 |
+
"learning_rate": 9.90051298775959e-06,
|
| 366 |
+
"loss": 0.8691,
|
| 367 |
+
"step": 51
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.8253968253968254,
|
| 371 |
+
"grad_norm": 0.3924890188917555,
|
| 372 |
+
"learning_rate": 9.88919395704952e-06,
|
| 373 |
+
"loss": 0.8259,
|
| 374 |
+
"step": 52
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.8412698412698413,
|
| 378 |
+
"grad_norm": 0.5584015479821739,
|
| 379 |
+
"learning_rate": 9.877272422691583e-06,
|
| 380 |
+
"loss": 0.9318,
|
| 381 |
+
"step": 53
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.8571428571428571,
|
| 385 |
+
"grad_norm": 0.5472693893320031,
|
| 386 |
+
"learning_rate": 9.864749853796642e-06,
|
| 387 |
+
"loss": 0.7983,
|
| 388 |
+
"step": 54
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.873015873015873,
|
| 392 |
+
"grad_norm": 0.5011856989250408,
|
| 393 |
+
"learning_rate": 9.85162779354206e-06,
|
| 394 |
+
"loss": 0.7289,
|
| 395 |
+
"step": 55
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.8888888888888888,
|
| 399 |
+
"grad_norm": 0.48176520075987733,
|
| 400 |
+
"learning_rate": 9.837907858981536e-06,
|
| 401 |
+
"loss": 0.8795,
|
| 402 |
+
"step": 56
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.9047619047619048,
|
| 406 |
+
"grad_norm": 0.4693619944653085,
|
| 407 |
+
"learning_rate": 9.823591740845831e-06,
|
| 408 |
+
"loss": 0.8625,
|
| 409 |
+
"step": 57
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.9206349206349206,
|
| 413 |
+
"grad_norm": 0.5158078748351012,
|
| 414 |
+
"learning_rate": 9.808681203334416e-06,
|
| 415 |
+
"loss": 0.7975,
|
| 416 |
+
"step": 58
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.9365079365079365,
|
| 420 |
+
"grad_norm": 0.467299048377056,
|
| 421 |
+
"learning_rate": 9.793178083898073e-06,
|
| 422 |
+
"loss": 0.878,
|
| 423 |
+
"step": 59
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.9523809523809523,
|
| 427 |
+
"grad_norm": 0.4360100853426926,
|
| 428 |
+
"learning_rate": 9.777084293012448e-06,
|
| 429 |
+
"loss": 0.842,
|
| 430 |
+
"step": 60
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.9682539682539683,
|
| 434 |
+
"grad_norm": 0.4999196363033725,
|
| 435 |
+
"learning_rate": 9.760401813942641e-06,
|
| 436 |
+
"loss": 0.7661,
|
| 437 |
+
"step": 61
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.9841269841269841,
|
| 441 |
+
"grad_norm": 0.49451715958225617,
|
| 442 |
+
"learning_rate": 9.743132702498785e-06,
|
| 443 |
+
"loss": 0.8685,
|
| 444 |
+
"step": 62
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 1.0,
|
| 448 |
+
"grad_norm": 0.51449429417728,
|
| 449 |
+
"learning_rate": 9.725279086782719e-06,
|
| 450 |
+
"loss": 0.7676,
|
| 451 |
+
"step": 63
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 1.0158730158730158,
|
| 455 |
+
"grad_norm": 0.5392465569053122,
|
| 456 |
+
"learning_rate": 9.706843166925733e-06,
|
| 457 |
+
"loss": 0.7978,
|
| 458 |
+
"step": 64
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 1.0317460317460316,
|
| 462 |
+
"grad_norm": 0.49426185655546884,
|
| 463 |
+
"learning_rate": 9.687827214817433e-06,
|
| 464 |
+
"loss": 0.8264,
|
| 465 |
+
"step": 65
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 1.0476190476190477,
|
| 469 |
+
"grad_norm": 0.5050909892528982,
|
| 470 |
+
"learning_rate": 9.668233573825794e-06,
|
| 471 |
+
"loss": 0.8898,
|
| 472 |
+
"step": 66
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 1.0634920634920635,
|
| 476 |
+
"grad_norm": 0.45134127922296613,
|
| 477 |
+
"learning_rate": 9.64806465850836e-06,
|
| 478 |
+
"loss": 0.7317,
|
| 479 |
+
"step": 67
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 1.0793650793650793,
|
| 483 |
+
"grad_norm": 0.5413266326970981,
|
| 484 |
+
"learning_rate": 9.62732295431471e-06,
|
| 485 |
+
"loss": 0.7307,
|
| 486 |
+
"step": 68
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 1.0952380952380953,
|
| 490 |
+
"grad_norm": 0.4781316290575908,
|
| 491 |
+
"learning_rate": 9.606011017280166e-06,
|
| 492 |
+
"loss": 0.8977,
|
| 493 |
+
"step": 69
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 1.1111111111111112,
|
| 497 |
+
"grad_norm": 0.5064140744904799,
|
| 498 |
+
"learning_rate": 9.5841314737108e-06,
|
| 499 |
+
"loss": 0.7648,
|
| 500 |
+
"step": 70
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 1.126984126984127,
|
| 504 |
+
"grad_norm": 0.5543523877170532,
|
| 505 |
+
"learning_rate": 9.56168701985981e-06,
|
| 506 |
+
"loss": 0.7995,
|
| 507 |
+
"step": 71
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 1.1428571428571428,
|
| 511 |
+
"grad_norm": 0.4891764300467825,
|
| 512 |
+
"learning_rate": 9.538680421595236e-06,
|
| 513 |
+
"loss": 0.8072,
|
| 514 |
+
"step": 72
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 1.1587301587301586,
|
| 518 |
+
"grad_norm": 0.48203192054287314,
|
| 519 |
+
"learning_rate": 9.515114514059127e-06,
|
| 520 |
+
"loss": 0.8128,
|
| 521 |
+
"step": 73
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 1.1746031746031746,
|
| 525 |
+
"grad_norm": 0.499915788005329,
|
| 526 |
+
"learning_rate": 9.490992201318165e-06,
|
| 527 |
+
"loss": 0.7876,
|
| 528 |
+
"step": 74
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 1.1904761904761905,
|
| 532 |
+
"grad_norm": 0.43129451868532453,
|
| 533 |
+
"learning_rate": 9.466316456005783e-06,
|
| 534 |
+
"loss": 0.7755,
|
| 535 |
+
"step": 75
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 1.2063492063492063,
|
| 539 |
+
"grad_norm": 0.49436944947590167,
|
| 540 |
+
"learning_rate": 9.441090318955843e-06,
|
| 541 |
+
"loss": 0.7015,
|
| 542 |
+
"step": 76
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 1.2222222222222223,
|
| 546 |
+
"grad_norm": 0.5018080177691097,
|
| 547 |
+
"learning_rate": 9.415316898827923e-06,
|
| 548 |
+
"loss": 0.7346,
|
| 549 |
+
"step": 77
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 1.2380952380952381,
|
| 553 |
+
"grad_norm": 0.42117192002428844,
|
| 554 |
+
"learning_rate": 9.388999371724212e-06,
|
| 555 |
+
"loss": 0.8242,
|
| 556 |
+
"step": 78
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 1.253968253968254,
|
| 560 |
+
"grad_norm": 0.47397540901194374,
|
| 561 |
+
"learning_rate": 9.362140980798127e-06,
|
| 562 |
+
"loss": 0.8928,
|
| 563 |
+
"step": 79
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 1.2698412698412698,
|
| 567 |
+
"grad_norm": 0.48823131897505534,
|
| 568 |
+
"learning_rate": 9.334745035854646e-06,
|
| 569 |
+
"loss": 0.7581,
|
| 570 |
+
"step": 80
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 1.2857142857142856,
|
| 574 |
+
"grad_norm": 0.5170460810325518,
|
| 575 |
+
"learning_rate": 9.306814912942445e-06,
|
| 576 |
+
"loss": 0.8361,
|
| 577 |
+
"step": 81
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 1.3015873015873016,
|
| 581 |
+
"grad_norm": 0.41118521047488926,
|
| 582 |
+
"learning_rate": 9.278354053937848e-06,
|
| 583 |
+
"loss": 0.7794,
|
| 584 |
+
"step": 82
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 1.3174603174603174,
|
| 588 |
+
"grad_norm": 0.4827654705693697,
|
| 589 |
+
"learning_rate": 9.249365966120692e-06,
|
| 590 |
+
"loss": 0.8542,
|
| 591 |
+
"step": 83
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 1.3333333333333333,
|
| 595 |
+
"grad_norm": 0.45176873751511454,
|
| 596 |
+
"learning_rate": 9.219854221742106e-06,
|
| 597 |
+
"loss": 0.8101,
|
| 598 |
+
"step": 84
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 1.3492063492063493,
|
| 602 |
+
"grad_norm": 0.44526540495239475,
|
| 603 |
+
"learning_rate": 9.189822457584311e-06,
|
| 604 |
+
"loss": 0.7419,
|
| 605 |
+
"step": 85
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 1.3650793650793651,
|
| 609 |
+
"grad_norm": 0.41133066066087726,
|
| 610 |
+
"learning_rate": 9.159274374512444e-06,
|
| 611 |
+
"loss": 0.6576,
|
| 612 |
+
"step": 86
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 1.380952380952381,
|
| 616 |
+
"grad_norm": 0.4500027229237173,
|
| 617 |
+
"learning_rate": 9.128213737018493e-06,
|
| 618 |
+
"loss": 0.8058,
|
| 619 |
+
"step": 87
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 1.3968253968253967,
|
| 623 |
+
"grad_norm": 0.40834920107678924,
|
| 624 |
+
"learning_rate": 9.096644372757393e-06,
|
| 625 |
+
"loss": 0.8849,
|
| 626 |
+
"step": 88
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 1.4126984126984126,
|
| 630 |
+
"grad_norm": 0.5843795042717066,
|
| 631 |
+
"learning_rate": 9.064570172075349e-06,
|
| 632 |
+
"loss": 0.7969,
|
| 633 |
+
"step": 89
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 1.4285714285714286,
|
| 637 |
+
"grad_norm": 0.5139681695756663,
|
| 638 |
+
"learning_rate": 9.031995087530403e-06,
|
| 639 |
+
"loss": 0.7983,
|
| 640 |
+
"step": 90
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 1.4444444444444444,
|
| 644 |
+
"grad_norm": 0.47799160571848326,
|
| 645 |
+
"learning_rate": 8.99892313340537e-06,
|
| 646 |
+
"loss": 0.6612,
|
| 647 |
+
"step": 91
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 1.4603174603174602,
|
| 651 |
+
"grad_norm": 0.48090290795792257,
|
| 652 |
+
"learning_rate": 8.96535838521314e-06,
|
| 653 |
+
"loss": 0.8026,
|
| 654 |
+
"step": 92
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 1.4761904761904763,
|
| 658 |
+
"grad_norm": 0.48955363216016506,
|
| 659 |
+
"learning_rate": 8.931304979194452e-06,
|
| 660 |
+
"loss": 0.8051,
|
| 661 |
+
"step": 93
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 1.492063492063492,
|
| 665 |
+
"grad_norm": 0.47949685756309185,
|
| 666 |
+
"learning_rate": 8.896767111808177e-06,
|
| 667 |
+
"loss": 0.7354,
|
| 668 |
+
"step": 94
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 1.507936507936508,
|
| 672 |
+
"grad_norm": 0.5732670061875946,
|
| 673 |
+
"learning_rate": 8.861749039214177e-06,
|
| 674 |
+
"loss": 0.9129,
|
| 675 |
+
"step": 95
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 1.5238095238095237,
|
| 679 |
+
"grad_norm": 0.48050508555262206,
|
| 680 |
+
"learning_rate": 8.826255076748823e-06,
|
| 681 |
+
"loss": 0.8445,
|
| 682 |
+
"step": 96
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 1.5396825396825395,
|
| 686 |
+
"grad_norm": 0.4329532952395629,
|
| 687 |
+
"learning_rate": 8.790289598393186e-06,
|
| 688 |
+
"loss": 0.7212,
|
| 689 |
+
"step": 97
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 1.5555555555555556,
|
| 693 |
+
"grad_norm": 0.522751486773223,
|
| 694 |
+
"learning_rate": 8.753857036234055e-06,
|
| 695 |
+
"loss": 0.8149,
|
| 696 |
+
"step": 98
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 1.5714285714285714,
|
| 700 |
+
"grad_norm": 0.4570961856172299,
|
| 701 |
+
"learning_rate": 8.716961879917734e-06,
|
| 702 |
+
"loss": 0.7365,
|
| 703 |
+
"step": 99
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 1.5873015873015874,
|
| 707 |
+
"grad_norm": 0.4363179134183329,
|
| 708 |
+
"learning_rate": 8.679608676096793e-06,
|
| 709 |
+
"loss": 0.8131,
|
| 710 |
+
"step": 100
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 1.6031746031746033,
|
| 714 |
+
"grad_norm": 0.4655541415571893,
|
| 715 |
+
"learning_rate": 8.641802027869774e-06,
|
| 716 |
+
"loss": 0.7946,
|
| 717 |
+
"step": 101
|
| 718 |
+
},
|
| 719 |
+
{
|
| 720 |
+
"epoch": 1.619047619047619,
|
| 721 |
+
"grad_norm": 0.5743139418639736,
|
| 722 |
+
"learning_rate": 8.603546594213935e-06,
|
| 723 |
+
"loss": 0.8574,
|
| 724 |
+
"step": 102
|
| 725 |
+
},
|
| 726 |
+
{
|
| 727 |
+
"epoch": 1.6349206349206349,
|
| 728 |
+
"grad_norm": 0.5267570867681096,
|
| 729 |
+
"learning_rate": 8.564847089411128e-06,
|
| 730 |
+
"loss": 0.8286,
|
| 731 |
+
"step": 103
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.6507936507936507,
|
| 735 |
+
"grad_norm": 0.40799736834923667,
|
| 736 |
+
"learning_rate": 8.525708282466839e-06,
|
| 737 |
+
"loss": 0.8412,
|
| 738 |
+
"step": 104
|
| 739 |
+
},
|
| 740 |
+
{
|
| 741 |
+
"epoch": 1.6666666666666665,
|
| 742 |
+
"grad_norm": 0.4236881481332967,
|
| 743 |
+
"learning_rate": 8.486134996522502e-06,
|
| 744 |
+
"loss": 0.8172,
|
| 745 |
+
"step": 105
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 1.6825396825396826,
|
| 749 |
+
"grad_norm": 0.5593679767726464,
|
| 750 |
+
"learning_rate": 8.446132108261136e-06,
|
| 751 |
+
"loss": 0.8058,
|
| 752 |
+
"step": 106
|
| 753 |
+
},
|
| 754 |
+
{
|
| 755 |
+
"epoch": 1.6984126984126984,
|
| 756 |
+
"grad_norm": 0.5031166228419733,
|
| 757 |
+
"learning_rate": 8.405704547306379e-06,
|
| 758 |
+
"loss": 0.8031,
|
| 759 |
+
"step": 107
|
| 760 |
+
},
|
| 761 |
+
{
|
| 762 |
+
"epoch": 1.7142857142857144,
|
| 763 |
+
"grad_norm": 0.45322610730579044,
|
| 764 |
+
"learning_rate": 8.364857295615006e-06,
|
| 765 |
+
"loss": 0.8903,
|
| 766 |
+
"step": 108
|
| 767 |
+
},
|
| 768 |
+
{
|
| 769 |
+
"epoch": 1.7301587301587302,
|
| 770 |
+
"grad_norm": 0.5335556769284883,
|
| 771 |
+
"learning_rate": 8.323595386862985e-06,
|
| 772 |
+
"loss": 0.7925,
|
| 773 |
+
"step": 109
|
| 774 |
+
},
|
| 775 |
+
{
|
| 776 |
+
"epoch": 1.746031746031746,
|
| 777 |
+
"grad_norm": 0.4699718024263939,
|
| 778 |
+
"learning_rate": 8.281923905825188e-06,
|
| 779 |
+
"loss": 0.7664,
|
| 780 |
+
"step": 110
|
| 781 |
+
},
|
| 782 |
+
{
|
| 783 |
+
"epoch": 1.7619047619047619,
|
| 784 |
+
"grad_norm": 0.47207237316096745,
|
| 785 |
+
"learning_rate": 8.23984798774876e-06,
|
| 786 |
+
"loss": 0.7347,
|
| 787 |
+
"step": 111
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"epoch": 1.7777777777777777,
|
| 791 |
+
"grad_norm": 0.4532509556288616,
|
| 792 |
+
"learning_rate": 8.197372817720314e-06,
|
| 793 |
+
"loss": 0.7369,
|
| 794 |
+
"step": 112
|
| 795 |
+
},
|
| 796 |
+
{
|
| 797 |
+
"epoch": 1.7936507936507935,
|
| 798 |
+
"grad_norm": 0.5443221798521994,
|
| 799 |
+
"learning_rate": 8.154503630026955e-06,
|
| 800 |
+
"loss": 0.7261,
|
| 801 |
+
"step": 113
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 1.8095238095238095,
|
| 805 |
+
"grad_norm": 0.4456098920838456,
|
| 806 |
+
"learning_rate": 8.111245707511253e-06,
|
| 807 |
+
"loss": 0.7194,
|
| 808 |
+
"step": 114
|
| 809 |
+
},
|
| 810 |
+
{
|
| 811 |
+
"epoch": 1.8253968253968254,
|
| 812 |
+
"grad_norm": 0.4159654938486175,
|
| 813 |
+
"learning_rate": 8.067604380920228e-06,
|
| 814 |
+
"loss": 0.7945,
|
| 815 |
+
"step": 115
|
| 816 |
+
},
|
| 817 |
+
{
|
| 818 |
+
"epoch": 1.8412698412698414,
|
| 819 |
+
"grad_norm": 0.4706342532274064,
|
| 820 |
+
"learning_rate": 8.023585028248435e-06,
|
| 821 |
+
"loss": 0.8487,
|
| 822 |
+
"step": 116
|
| 823 |
+
},
|
| 824 |
+
{
|
| 825 |
+
"epoch": 1.8571428571428572,
|
| 826 |
+
"grad_norm": 0.5701232470412769,
|
| 827 |
+
"learning_rate": 7.979193074075216e-06,
|
| 828 |
+
"loss": 0.8887,
|
| 829 |
+
"step": 117
|
| 830 |
+
},
|
| 831 |
+
{
|
| 832 |
+
"epoch": 1.873015873015873,
|
| 833 |
+
"grad_norm": 0.530430629054239,
|
| 834 |
+
"learning_rate": 7.934433988896233e-06,
|
| 835 |
+
"loss": 0.6534,
|
| 836 |
+
"step": 118
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.8888888888888888,
|
| 840 |
+
"grad_norm": 0.48414840419963984,
|
| 841 |
+
"learning_rate": 7.889313288449323e-06,
|
| 842 |
+
"loss": 0.8214,
|
| 843 |
+
"step": 119
|
| 844 |
+
},
|
| 845 |
+
{
|
| 846 |
+
"epoch": 1.9047619047619047,
|
| 847 |
+
"grad_norm": 0.4200926363513126,
|
| 848 |
+
"learning_rate": 7.843836533034784e-06,
|
| 849 |
+
"loss": 0.7614,
|
| 850 |
+
"step": 120
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 1.9206349206349205,
|
| 854 |
+
"grad_norm": 0.4941849127950555,
|
| 855 |
+
"learning_rate": 7.798009326830167e-06,
|
| 856 |
+
"loss": 0.7996,
|
| 857 |
+
"step": 121
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 1.9365079365079365,
|
| 861 |
+
"grad_norm": 0.41647477043231534,
|
| 862 |
+
"learning_rate": 7.751837317199673e-06,
|
| 863 |
+
"loss": 0.867,
|
| 864 |
+
"step": 122
|
| 865 |
+
},
|
| 866 |
+
{
|
| 867 |
+
"epoch": 1.9523809523809523,
|
| 868 |
+
"grad_norm": 0.4462896414872465,
|
| 869 |
+
"learning_rate": 7.705326193998207e-06,
|
| 870 |
+
"loss": 0.7547,
|
| 871 |
+
"step": 123
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 1.9682539682539684,
|
| 875 |
+
"grad_norm": 0.46366747032871125,
|
| 876 |
+
"learning_rate": 7.658481688870218e-06,
|
| 877 |
+
"loss": 0.7582,
|
| 878 |
+
"step": 124
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 1.9841269841269842,
|
| 882 |
+
"grad_norm": 0.4714130206121814,
|
| 883 |
+
"learning_rate": 7.611309574543373e-06,
|
| 884 |
+
"loss": 0.7606,
|
| 885 |
+
"step": 125
|
| 886 |
+
},
|
| 887 |
+
{
|
| 888 |
+
"epoch": 2.0,
|
| 889 |
+
"grad_norm": 0.46690869317456135,
|
| 890 |
+
"learning_rate": 7.563815664117173e-06,
|
| 891 |
+
"loss": 0.9121,
|
| 892 |
+
"step": 126
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 2.015873015873016,
|
| 896 |
+
"grad_norm": 0.8060769356732992,
|
| 897 |
+
"learning_rate": 7.5160058103465985e-06,
|
| 898 |
+
"loss": 0.7122,
|
| 899 |
+
"step": 127
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 2.0317460317460316,
|
| 903 |
+
"grad_norm": 0.5953210710991091,
|
| 904 |
+
"learning_rate": 7.467885904920864e-06,
|
| 905 |
+
"loss": 0.7567,
|
| 906 |
+
"step": 128
|
| 907 |
+
},
|
| 908 |
+
{
|
| 909 |
+
"epoch": 2.0476190476190474,
|
| 910 |
+
"grad_norm": 0.7560871980312371,
|
| 911 |
+
"learning_rate": 7.419461877737373e-06,
|
| 912 |
+
"loss": 0.8318,
|
| 913 |
+
"step": 129
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 2.0634920634920633,
|
| 917 |
+
"grad_norm": 2.1011598702400667,
|
| 918 |
+
"learning_rate": 7.370739696170971e-06,
|
| 919 |
+
"loss": 0.7428,
|
| 920 |
+
"step": 130
|
| 921 |
+
},
|
| 922 |
+
{
|
| 923 |
+
"epoch": 2.0793650793650795,
|
| 924 |
+
"grad_norm": 0.940827314326734,
|
| 925 |
+
"learning_rate": 7.321725364338566e-06,
|
| 926 |
+
"loss": 0.6161,
|
| 927 |
+
"step": 131
|
| 928 |
+
},
|
| 929 |
+
{
|
| 930 |
+
"epoch": 2.0952380952380953,
|
| 931 |
+
"grad_norm": 0.5381465457966281,
|
| 932 |
+
"learning_rate": 7.272424922359246e-06,
|
| 933 |
+
"loss": 0.6432,
|
| 934 |
+
"step": 132
|
| 935 |
+
},
|
| 936 |
+
{
|
| 937 |
+
"epoch": 2.111111111111111,
|
| 938 |
+
"grad_norm": 0.7212158226191104,
|
| 939 |
+
"learning_rate": 7.222844445609931e-06,
|
| 940 |
+
"loss": 0.7817,
|
| 941 |
+
"step": 133
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 2.126984126984127,
|
| 945 |
+
"grad_norm": 0.6031927565028607,
|
| 946 |
+
"learning_rate": 7.172990043976703e-06,
|
| 947 |
+
"loss": 0.7291,
|
| 948 |
+
"step": 134
|
| 949 |
+
},
|
| 950 |
+
{
|
| 951 |
+
"epoch": 2.142857142857143,
|
| 952 |
+
"grad_norm": 0.5554913039306149,
|
| 953 |
+
"learning_rate": 7.122867861101868e-06,
|
| 954 |
+
"loss": 0.7928,
|
| 955 |
+
"step": 135
|
| 956 |
+
},
|
| 957 |
+
{
|
| 958 |
+
"epoch": 2.1587301587301586,
|
| 959 |
+
"grad_norm": 0.4440614086169425,
|
| 960 |
+
"learning_rate": 7.072484073626872e-06,
|
| 961 |
+
"loss": 0.6864,
|
| 962 |
+
"step": 136
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"epoch": 2.1746031746031744,
|
| 966 |
+
"grad_norm": 2.875286251781212,
|
| 967 |
+
"learning_rate": 7.021844890431136e-06,
|
| 968 |
+
"loss": 0.7627,
|
| 969 |
+
"step": 137
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 2.1904761904761907,
|
| 973 |
+
"grad_norm": 0.7248206082063566,
|
| 974 |
+
"learning_rate": 6.970956551866925e-06,
|
| 975 |
+
"loss": 0.728,
|
| 976 |
+
"step": 138
|
| 977 |
+
},
|
| 978 |
+
{
|
| 979 |
+
"epoch": 2.2063492063492065,
|
| 980 |
+
"grad_norm": 0.5344769870855947,
|
| 981 |
+
"learning_rate": 6.9198253289903515e-06,
|
| 982 |
+
"loss": 0.6621,
|
| 983 |
+
"step": 139
|
| 984 |
+
},
|
| 985 |
+
{
|
| 986 |
+
"epoch": 2.2222222222222223,
|
| 987 |
+
"grad_norm": 0.5334409779130068,
|
| 988 |
+
"learning_rate": 6.868457522788561e-06,
|
| 989 |
+
"loss": 0.7351,
|
| 990 |
+
"step": 140
|
| 991 |
+
},
|
| 992 |
+
{
|
| 993 |
+
"epoch": 2.238095238095238,
|
| 994 |
+
"grad_norm": 0.4791675678917909,
|
| 995 |
+
"learning_rate": 6.816859463403271e-06,
|
| 996 |
+
"loss": 0.6568,
|
| 997 |
+
"step": 141
|
| 998 |
+
},
|
| 999 |
+
{
|
| 1000 |
+
"epoch": 2.253968253968254,
|
| 1001 |
+
"grad_norm": 0.5667962259074942,
|
| 1002 |
+
"learning_rate": 6.765037509350685e-06,
|
| 1003 |
+
"loss": 0.758,
|
| 1004 |
+
"step": 142
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"epoch": 2.2698412698412698,
|
| 1008 |
+
"grad_norm": 0.523154654898243,
|
| 1009 |
+
"learning_rate": 6.7129980467379265e-06,
|
| 1010 |
+
"loss": 0.6657,
|
| 1011 |
+
"step": 143
|
| 1012 |
+
},
|
| 1013 |
+
{
|
| 1014 |
+
"epoch": 2.2857142857142856,
|
| 1015 |
+
"grad_norm": 0.45239550513741295,
|
| 1016 |
+
"learning_rate": 6.660747488476066e-06,
|
| 1017 |
+
"loss": 0.6615,
|
| 1018 |
+
"step": 144
|
| 1019 |
+
},
|
| 1020 |
+
{
|
| 1021 |
+
"epoch": 2.3015873015873014,
|
| 1022 |
+
"grad_norm": 0.6580127713752147,
|
| 1023 |
+
"learning_rate": 6.608292273489851e-06,
|
| 1024 |
+
"loss": 0.6112,
|
| 1025 |
+
"step": 145
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 2.317460317460317,
|
| 1029 |
+
"grad_norm": 0.6033248382665617,
|
| 1030 |
+
"learning_rate": 6.555638865924221e-06,
|
| 1031 |
+
"loss": 0.7033,
|
| 1032 |
+
"step": 146
|
| 1033 |
+
},
|
| 1034 |
+
{
|
| 1035 |
+
"epoch": 2.3333333333333335,
|
| 1036 |
+
"grad_norm": 0.5578461067365529,
|
| 1037 |
+
"learning_rate": 6.502793754347721e-06,
|
| 1038 |
+
"loss": 0.7578,
|
| 1039 |
+
"step": 147
|
| 1040 |
+
},
|
| 1041 |
+
{
|
| 1042 |
+
"epoch": 2.3492063492063493,
|
| 1043 |
+
"grad_norm": 0.5451762654132818,
|
| 1044 |
+
"learning_rate": 6.449763450952912e-06,
|
| 1045 |
+
"loss": 0.6863,
|
| 1046 |
+
"step": 148
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 2.365079365079365,
|
| 1050 |
+
"grad_norm": 0.4667833185680937,
|
| 1051 |
+
"learning_rate": 6.396554490753848e-06,
|
| 1052 |
+
"loss": 0.6825,
|
| 1053 |
+
"step": 149
|
| 1054 |
+
},
|
| 1055 |
+
{
|
| 1056 |
+
"epoch": 2.380952380952381,
|
| 1057 |
+
"grad_norm": 0.4209933154088852,
|
| 1058 |
+
"learning_rate": 6.343173430780769e-06,
|
| 1059 |
+
"loss": 0.836,
|
| 1060 |
+
"step": 150
|
| 1061 |
+
},
|
| 1062 |
+
{
|
| 1063 |
+
"epoch": 2.3968253968253967,
|
| 1064 |
+
"grad_norm": 0.46876037251704294,
|
| 1065 |
+
"learning_rate": 6.289626849272062e-06,
|
| 1066 |
+
"loss": 0.7981,
|
| 1067 |
+
"step": 151
|
| 1068 |
+
},
|
| 1069 |
+
{
|
| 1070 |
+
"epoch": 2.4126984126984126,
|
| 1071 |
+
"grad_norm": 0.47367833829704725,
|
| 1072 |
+
"learning_rate": 6.2359213448636104e-06,
|
| 1073 |
+
"loss": 0.751,
|
| 1074 |
+
"step": 152
|
| 1075 |
+
},
|
| 1076 |
+
{
|
| 1077 |
+
"epoch": 2.4285714285714284,
|
| 1078 |
+
"grad_norm": 0.43459439089398605,
|
| 1079 |
+
"learning_rate": 6.182063535775634e-06,
|
| 1080 |
+
"loss": 0.7654,
|
| 1081 |
+
"step": 153
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 2.4444444444444446,
|
| 1085 |
+
"grad_norm": 0.39767398947957067,
|
| 1086 |
+
"learning_rate": 6.1280600589971225e-06,
|
| 1087 |
+
"loss": 0.7896,
|
| 1088 |
+
"step": 154
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 2.4603174603174605,
|
| 1092 |
+
"grad_norm": 0.4231324131775063,
|
| 1093 |
+
"learning_rate": 6.073917569467934e-06,
|
| 1094 |
+
"loss": 0.8051,
|
| 1095 |
+
"step": 155
|
| 1096 |
+
},
|
| 1097 |
+
{
|
| 1098 |
+
"epoch": 2.4761904761904763,
|
| 1099 |
+
"grad_norm": 0.3983830637612639,
|
| 1100 |
+
"learning_rate": 6.0196427392587085e-06,
|
| 1101 |
+
"loss": 0.7038,
|
| 1102 |
+
"step": 156
|
| 1103 |
+
},
|
| 1104 |
+
{
|
| 1105 |
+
"epoch": 2.492063492063492,
|
| 1106 |
+
"grad_norm": 0.4585701856768339,
|
| 1107 |
+
"learning_rate": 5.96524225674865e-06,
|
| 1108 |
+
"loss": 0.7422,
|
| 1109 |
+
"step": 157
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 2.507936507936508,
|
| 1113 |
+
"grad_norm": 0.4299692751487169,
|
| 1114 |
+
"learning_rate": 5.9107228258013085e-06,
|
| 1115 |
+
"loss": 0.7053,
|
| 1116 |
+
"step": 158
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 2.5238095238095237,
|
| 1120 |
+
"grad_norm": 0.42827944956580943,
|
| 1121 |
+
"learning_rate": 5.856091164938451e-06,
|
| 1122 |
+
"loss": 0.6523,
|
| 1123 |
+
"step": 159
|
| 1124 |
+
},
|
| 1125 |
+
{
|
| 1126 |
+
"epoch": 2.5396825396825395,
|
| 1127 |
+
"grad_norm": 0.4098750307712162,
|
| 1128 |
+
"learning_rate": 5.801354006512127e-06,
|
| 1129 |
+
"loss": 0.6895,
|
| 1130 |
+
"step": 160
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 2.5555555555555554,
|
| 1134 |
+
"grad_norm": 0.45624383692077836,
|
| 1135 |
+
"learning_rate": 5.746518095875033e-06,
|
| 1136 |
+
"loss": 0.6973,
|
| 1137 |
+
"step": 161
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 2.571428571428571,
|
| 1141 |
+
"grad_norm": 0.40961695420487504,
|
| 1142 |
+
"learning_rate": 5.6915901905492586e-06,
|
| 1143 |
+
"loss": 0.6285,
|
| 1144 |
+
"step": 162
|
| 1145 |
+
},
|
| 1146 |
+
{
|
| 1147 |
+
"epoch": 2.5873015873015874,
|
| 1148 |
+
"grad_norm": 0.5946218628280344,
|
| 1149 |
+
"learning_rate": 5.6365770593935665e-06,
|
| 1150 |
+
"loss": 0.5907,
|
| 1151 |
+
"step": 163
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 2.6031746031746033,
|
| 1155 |
+
"grad_norm": 0.5401440035651196,
|
| 1156 |
+
"learning_rate": 5.581485481769231e-06,
|
| 1157 |
+
"loss": 0.7181,
|
| 1158 |
+
"step": 164
|
| 1159 |
+
},
|
| 1160 |
+
{
|
| 1161 |
+
"epoch": 2.619047619047619,
|
| 1162 |
+
"grad_norm": 0.4378876946579892,
|
| 1163 |
+
"learning_rate": 5.526322246704628e-06,
|
| 1164 |
+
"loss": 0.7978,
|
| 1165 |
+
"step": 165
|
| 1166 |
+
},
|
| 1167 |
+
{
|
| 1168 |
+
"epoch": 2.634920634920635,
|
| 1169 |
+
"grad_norm": 0.40853074862176036,
|
| 1170 |
+
"learning_rate": 5.471094152058592e-06,
|
| 1171 |
+
"loss": 0.681,
|
| 1172 |
+
"step": 166
|
| 1173 |
+
},
|
| 1174 |
+
{
|
| 1175 |
+
"epoch": 2.6507936507936507,
|
| 1176 |
+
"grad_norm": 0.46389651051528763,
|
| 1177 |
+
"learning_rate": 5.415808003682717e-06,
|
| 1178 |
+
"loss": 0.7308,
|
| 1179 |
+
"step": 167
|
| 1180 |
+
},
|
| 1181 |
+
{
|
| 1182 |
+
"epoch": 2.6666666666666665,
|
| 1183 |
+
"grad_norm": 0.42992484133372394,
|
| 1184 |
+
"learning_rate": 5.360470614582661e-06,
|
| 1185 |
+
"loss": 0.7136,
|
| 1186 |
+
"step": 168
|
| 1187 |
+
},
|
| 1188 |
+
{
|
| 1189 |
+
"epoch": 2.682539682539683,
|
| 1190 |
+
"grad_norm": 0.557870344379466,
|
| 1191 |
+
"learning_rate": 5.305088804078559e-06,
|
| 1192 |
+
"loss": 0.7333,
|
| 1193 |
+
"step": 169
|
| 1194 |
+
},
|
| 1195 |
+
{
|
| 1196 |
+
"epoch": 2.6984126984126986,
|
| 1197 |
+
"grad_norm": 0.45453618737081114,
|
| 1198 |
+
"learning_rate": 5.249669396964665e-06,
|
| 1199 |
+
"loss": 0.6349,
|
| 1200 |
+
"step": 170
|
| 1201 |
+
},
|
| 1202 |
+
{
|
| 1203 |
+
"epoch": 2.7142857142857144,
|
| 1204 |
+
"grad_norm": 0.4511080452383348,
|
| 1205 |
+
"learning_rate": 5.1942192226683385e-06,
|
| 1206 |
+
"loss": 0.776,
|
| 1207 |
+
"step": 171
|
| 1208 |
+
},
|
| 1209 |
+
{
|
| 1210 |
+
"epoch": 2.7301587301587302,
|
| 1211 |
+
"grad_norm": 0.4900595891663431,
|
| 1212 |
+
"learning_rate": 5.138745114408427e-06,
|
| 1213 |
+
"loss": 0.5998,
|
| 1214 |
+
"step": 172
|
| 1215 |
+
},
|
| 1216 |
+
{
|
| 1217 |
+
"epoch": 2.746031746031746,
|
| 1218 |
+
"grad_norm": 0.46419645481002475,
|
| 1219 |
+
"learning_rate": 5.083253908353193e-06,
|
| 1220 |
+
"loss": 0.6676,
|
| 1221 |
+
"step": 173
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"epoch": 2.761904761904762,
|
| 1225 |
+
"grad_norm": 0.4905121964865482,
|
| 1226 |
+
"learning_rate": 5.0277524427778986e-06,
|
| 1227 |
+
"loss": 0.7831,
|
| 1228 |
+
"step": 174
|
| 1229 |
+
},
|
| 1230 |
+
{
|
| 1231 |
+
"epoch": 2.7777777777777777,
|
| 1232 |
+
"grad_norm": 0.4525848803424086,
|
| 1233 |
+
"learning_rate": 4.972247557222102e-06,
|
| 1234 |
+
"loss": 0.7164,
|
| 1235 |
+
"step": 175
|
| 1236 |
+
},
|
| 1237 |
+
{
|
| 1238 |
+
"epoch": 2.7936507936507935,
|
| 1239 |
+
"grad_norm": 0.5368330661361714,
|
| 1240 |
+
"learning_rate": 4.916746091646808e-06,
|
| 1241 |
+
"loss": 0.6805,
|
| 1242 |
+
"step": 176
|
| 1243 |
+
},
|
| 1244 |
+
{
|
| 1245 |
+
"epoch": 2.8095238095238093,
|
| 1246 |
+
"grad_norm": 0.4224136348005534,
|
| 1247 |
+
"learning_rate": 4.8612548855915755e-06,
|
| 1248 |
+
"loss": 0.724,
|
| 1249 |
+
"step": 177
|
| 1250 |
+
},
|
| 1251 |
+
{
|
| 1252 |
+
"epoch": 2.825396825396825,
|
| 1253 |
+
"grad_norm": 0.4869146817578471,
|
| 1254 |
+
"learning_rate": 4.805780777331662e-06,
|
| 1255 |
+
"loss": 0.7446,
|
| 1256 |
+
"step": 178
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 2.8412698412698414,
|
| 1260 |
+
"grad_norm": 0.4511553539717499,
|
| 1261 |
+
"learning_rate": 4.750330603035336e-06,
|
| 1262 |
+
"loss": 0.7124,
|
| 1263 |
+
"step": 179
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"epoch": 2.857142857142857,
|
| 1267 |
+
"grad_norm": 0.45286461188582156,
|
| 1268 |
+
"learning_rate": 4.694911195921443e-06,
|
| 1269 |
+
"loss": 0.7252,
|
| 1270 |
+
"step": 180
|
| 1271 |
+
},
|
| 1272 |
+
{
|
| 1273 |
+
"epoch": 2.873015873015873,
|
| 1274 |
+
"grad_norm": 0.4293856541441545,
|
| 1275 |
+
"learning_rate": 4.6395293854173395e-06,
|
| 1276 |
+
"loss": 0.6053,
|
| 1277 |
+
"step": 181
|
| 1278 |
+
},
|
| 1279 |
+
{
|
| 1280 |
+
"epoch": 2.888888888888889,
|
| 1281 |
+
"grad_norm": 0.5836445711298119,
|
| 1282 |
+
"learning_rate": 4.584191996317285e-06,
|
| 1283 |
+
"loss": 0.6828,
|
| 1284 |
+
"step": 182
|
| 1285 |
+
},
|
| 1286 |
+
{
|
| 1287 |
+
"epoch": 2.9047619047619047,
|
| 1288 |
+
"grad_norm": 0.4710990644177235,
|
| 1289 |
+
"learning_rate": 4.528905847941411e-06,
|
| 1290 |
+
"loss": 0.8414,
|
| 1291 |
+
"step": 183
|
| 1292 |
+
},
|
| 1293 |
+
{
|
| 1294 |
+
"epoch": 2.9206349206349205,
|
| 1295 |
+
"grad_norm": 0.4770727006845428,
|
| 1296 |
+
"learning_rate": 4.473677753295375e-06,
|
| 1297 |
+
"loss": 0.6592,
|
| 1298 |
+
"step": 184
|
| 1299 |
+
},
|
| 1300 |
+
{
|
| 1301 |
+
"epoch": 2.9365079365079367,
|
| 1302 |
+
"grad_norm": 0.4087196329651188,
|
| 1303 |
+
"learning_rate": 4.418514518230769e-06,
|
| 1304 |
+
"loss": 0.7122,
|
| 1305 |
+
"step": 185
|
| 1306 |
+
},
|
| 1307 |
+
{
|
| 1308 |
+
"epoch": 2.9523809523809526,
|
| 1309 |
+
"grad_norm": 0.48530953017287554,
|
| 1310 |
+
"learning_rate": 4.363422940606435e-06,
|
| 1311 |
+
"loss": 0.7454,
|
| 1312 |
+
"step": 186
|
| 1313 |
+
},
|
| 1314 |
+
{
|
| 1315 |
+
"epoch": 2.9682539682539684,
|
| 1316 |
+
"grad_norm": 0.4900655757859956,
|
| 1317 |
+
"learning_rate": 4.308409809450742e-06,
|
| 1318 |
+
"loss": 0.7621,
|
| 1319 |
+
"step": 187
|
| 1320 |
+
},
|
| 1321 |
+
{
|
| 1322 |
+
"epoch": 2.984126984126984,
|
| 1323 |
+
"grad_norm": 0.49992358594135816,
|
| 1324 |
+
"learning_rate": 4.253481904124968e-06,
|
| 1325 |
+
"loss": 0.7331,
|
| 1326 |
+
"step": 188
|
| 1327 |
+
},
|
| 1328 |
+
{
|
| 1329 |
+
"epoch": 3.0,
|
| 1330 |
+
"grad_norm": 0.4385555753140365,
|
| 1331 |
+
"learning_rate": 4.198645993487872e-06,
|
| 1332 |
+
"loss": 0.604,
|
| 1333 |
+
"step": 189
|
| 1334 |
+
},
|
| 1335 |
+
{
|
| 1336 |
+
"epoch": 3.015873015873016,
|
| 1337 |
+
"grad_norm": 0.8387072939566776,
|
| 1338 |
+
"learning_rate": 4.143908835061551e-06,
|
| 1339 |
+
"loss": 0.6844,
|
| 1340 |
+
"step": 190
|
| 1341 |
+
},
|
| 1342 |
+
{
|
| 1343 |
+
"epoch": 3.0317460317460316,
|
| 1344 |
+
"grad_norm": 0.5067553850596358,
|
| 1345 |
+
"learning_rate": 4.089277174198694e-06,
|
| 1346 |
+
"loss": 0.7,
|
| 1347 |
+
"step": 191
|
| 1348 |
+
},
|
| 1349 |
+
{
|
| 1350 |
+
"epoch": 3.0476190476190474,
|
| 1351 |
+
"grad_norm": 0.8011190701762613,
|
| 1352 |
+
"learning_rate": 4.0347577432513515e-06,
|
| 1353 |
+
"loss": 0.743,
|
| 1354 |
+
"step": 192
|
| 1355 |
+
},
|
| 1356 |
+
{
|
| 1357 |
+
"epoch": 3.0634920634920633,
|
| 1358 |
+
"grad_norm": 0.5036174850503073,
|
| 1359 |
+
"learning_rate": 3.980357260741293e-06,
|
| 1360 |
+
"loss": 0.6811,
|
| 1361 |
+
"step": 193
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 3.0793650793650795,
|
| 1365 |
+
"grad_norm": 0.5463958175244662,
|
| 1366 |
+
"learning_rate": 3.926082430532067e-06,
|
| 1367 |
+
"loss": 0.6461,
|
| 1368 |
+
"step": 194
|
| 1369 |
+
},
|
| 1370 |
+
{
|
| 1371 |
+
"epoch": 3.0952380952380953,
|
| 1372 |
+
"grad_norm": 0.6797072746043096,
|
| 1373 |
+
"learning_rate": 3.87193994100288e-06,
|
| 1374 |
+
"loss": 0.6076,
|
| 1375 |
+
"step": 195
|
| 1376 |
+
},
|
| 1377 |
+
{
|
| 1378 |
+
"epoch": 3.111111111111111,
|
| 1379 |
+
"grad_norm": 0.6595821505262829,
|
| 1380 |
+
"learning_rate": 3.817936464224367e-06,
|
| 1381 |
+
"loss": 0.6729,
|
| 1382 |
+
"step": 196
|
| 1383 |
+
},
|
| 1384 |
+
{
|
| 1385 |
+
"epoch": 3.126984126984127,
|
| 1386 |
+
"grad_norm": 0.6303031378119635,
|
| 1387 |
+
"learning_rate": 3.764078655136391e-06,
|
| 1388 |
+
"loss": 0.7462,
|
| 1389 |
+
"step": 197
|
| 1390 |
+
},
|
| 1391 |
+
{
|
| 1392 |
+
"epoch": 3.142857142857143,
|
| 1393 |
+
"grad_norm": 0.5867535234820467,
|
| 1394 |
+
"learning_rate": 3.7103731507279383e-06,
|
| 1395 |
+
"loss": 0.6994,
|
| 1396 |
+
"step": 198
|
| 1397 |
+
},
|
| 1398 |
+
{
|
| 1399 |
+
"epoch": 3.1587301587301586,
|
| 1400 |
+
"grad_norm": 0.744206344239702,
|
| 1401 |
+
"learning_rate": 3.656826569219233e-06,
|
| 1402 |
+
"loss": 0.6698,
|
| 1403 |
+
"step": 199
|
| 1404 |
+
},
|
| 1405 |
+
{
|
| 1406 |
+
"epoch": 3.1746031746031744,
|
| 1407 |
+
"grad_norm": 0.48008260540189834,
|
| 1408 |
+
"learning_rate": 3.603445509246154e-06,
|
| 1409 |
+
"loss": 0.641,
|
| 1410 |
+
"step": 200
|
| 1411 |
+
},
|
| 1412 |
+
{
|
| 1413 |
+
"epoch": 3.1904761904761907,
|
| 1414 |
+
"grad_norm": 0.4479357669097382,
|
| 1415 |
+
"learning_rate": 3.55023654904709e-06,
|
| 1416 |
+
"loss": 0.7151,
|
| 1417 |
+
"step": 201
|
| 1418 |
+
},
|
| 1419 |
+
{
|
| 1420 |
+
"epoch": 3.2063492063492065,
|
| 1421 |
+
"grad_norm": 0.4992070753022617,
|
| 1422 |
+
"learning_rate": 3.49720624565228e-06,
|
| 1423 |
+
"loss": 0.5512,
|
| 1424 |
+
"step": 202
|
| 1425 |
+
},
|
| 1426 |
+
{
|
| 1427 |
+
"epoch": 3.2222222222222223,
|
| 1428 |
+
"grad_norm": 0.48763610007436026,
|
| 1429 |
+
"learning_rate": 3.44436113407578e-06,
|
| 1430 |
+
"loss": 0.6851,
|
| 1431 |
+
"step": 203
|
| 1432 |
+
},
|
| 1433 |
+
{
|
| 1434 |
+
"epoch": 3.238095238095238,
|
| 1435 |
+
"grad_norm": 0.5828989750562972,
|
| 1436 |
+
"learning_rate": 3.3917077265101505e-06,
|
| 1437 |
+
"loss": 0.7484,
|
| 1438 |
+
"step": 204
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"epoch": 3.253968253968254,
|
| 1442 |
+
"grad_norm": 0.5181196242222144,
|
| 1443 |
+
"learning_rate": 3.3392525115239353e-06,
|
| 1444 |
+
"loss": 0.5734,
|
| 1445 |
+
"step": 205
|
| 1446 |
+
},
|
| 1447 |
+
{
|
| 1448 |
+
"epoch": 3.2698412698412698,
|
| 1449 |
+
"grad_norm": 0.47471736105285633,
|
| 1450 |
+
"learning_rate": 3.2870019532620744e-06,
|
| 1451 |
+
"loss": 0.7111,
|
| 1452 |
+
"step": 206
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 3.2857142857142856,
|
| 1456 |
+
"grad_norm": 0.44233225614290655,
|
| 1457 |
+
"learning_rate": 3.2349624906493164e-06,
|
| 1458 |
+
"loss": 0.6547,
|
| 1459 |
+
"step": 207
|
| 1460 |
+
},
|
| 1461 |
+
{
|
| 1462 |
+
"epoch": 3.3015873015873014,
|
| 1463 |
+
"grad_norm": 0.549412981718262,
|
| 1464 |
+
"learning_rate": 3.1831405365967315e-06,
|
| 1465 |
+
"loss": 0.5268,
|
| 1466 |
+
"step": 208
|
| 1467 |
+
},
|
| 1468 |
+
{
|
| 1469 |
+
"epoch": 3.317460317460317,
|
| 1470 |
+
"grad_norm": 0.4610400286542981,
|
| 1471 |
+
"learning_rate": 3.1315424772114404e-06,
|
| 1472 |
+
"loss": 0.7138,
|
| 1473 |
+
"step": 209
|
| 1474 |
+
},
|
| 1475 |
+
{
|
| 1476 |
+
"epoch": 3.3333333333333335,
|
| 1477 |
+
"grad_norm": 0.5912978101838896,
|
| 1478 |
+
"learning_rate": 3.0801746710096497e-06,
|
| 1479 |
+
"loss": 0.5523,
|
| 1480 |
+
"step": 210
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"epoch": 3.3492063492063493,
|
| 1484 |
+
"grad_norm": 0.5365898687480075,
|
| 1485 |
+
"learning_rate": 3.0290434481330746e-06,
|
| 1486 |
+
"loss": 0.586,
|
| 1487 |
+
"step": 211
|
| 1488 |
+
},
|
| 1489 |
+
{
|
| 1490 |
+
"epoch": 3.365079365079365,
|
| 1491 |
+
"grad_norm": 0.46259460742378206,
|
| 1492 |
+
"learning_rate": 2.978155109568864e-06,
|
| 1493 |
+
"loss": 0.6185,
|
| 1494 |
+
"step": 212
|
| 1495 |
+
},
|
| 1496 |
+
{
|
| 1497 |
+
"epoch": 3.380952380952381,
|
| 1498 |
+
"grad_norm": 0.5035201915907002,
|
| 1499 |
+
"learning_rate": 2.927515926373129e-06,
|
| 1500 |
+
"loss": 0.6636,
|
| 1501 |
+
"step": 213
|
| 1502 |
+
},
|
| 1503 |
+
{
|
| 1504 |
+
"epoch": 3.3968253968253967,
|
| 1505 |
+
"grad_norm": 0.45266495724596634,
|
| 1506 |
+
"learning_rate": 2.8771321388981334e-06,
|
| 1507 |
+
"loss": 0.6636,
|
| 1508 |
+
"step": 214
|
| 1509 |
+
},
|
| 1510 |
+
{
|
| 1511 |
+
"epoch": 3.4126984126984126,
|
| 1512 |
+
"grad_norm": 0.9391085088935888,
|
| 1513 |
+
"learning_rate": 2.8270099560232992e-06,
|
| 1514 |
+
"loss": 0.7087,
|
| 1515 |
+
"step": 215
|
| 1516 |
+
},
|
| 1517 |
+
{
|
| 1518 |
+
"epoch": 3.4285714285714284,
|
| 1519 |
+
"grad_norm": 0.4858344733348455,
|
| 1520 |
+
"learning_rate": 2.77715555439007e-06,
|
| 1521 |
+
"loss": 0.7505,
|
| 1522 |
+
"step": 216
|
| 1523 |
+
},
|
| 1524 |
+
{
|
| 1525 |
+
"epoch": 3.4444444444444446,
|
| 1526 |
+
"grad_norm": 0.43967486893271346,
|
| 1527 |
+
"learning_rate": 2.7275750776407568e-06,
|
| 1528 |
+
"loss": 0.6028,
|
| 1529 |
+
"step": 217
|
| 1530 |
+
},
|
| 1531 |
+
{
|
| 1532 |
+
"epoch": 3.4603174603174605,
|
| 1533 |
+
"grad_norm": 0.4717225483417324,
|
| 1534 |
+
"learning_rate": 2.6782746356614364e-06,
|
| 1535 |
+
"loss": 0.5843,
|
| 1536 |
+
"step": 218
|
| 1537 |
+
},
|
| 1538 |
+
{
|
| 1539 |
+
"epoch": 3.4761904761904763,
|
| 1540 |
+
"grad_norm": 0.5642208112077767,
|
| 1541 |
+
"learning_rate": 2.6292603038290306e-06,
|
| 1542 |
+
"loss": 0.7137,
|
| 1543 |
+
"step": 219
|
| 1544 |
+
},
|
| 1545 |
+
{
|
| 1546 |
+
"epoch": 3.492063492063492,
|
| 1547 |
+
"grad_norm": 0.49853539661605933,
|
| 1548 |
+
"learning_rate": 2.580538122262627e-06,
|
| 1549 |
+
"loss": 0.6527,
|
| 1550 |
+
"step": 220
|
| 1551 |
+
},
|
| 1552 |
+
{
|
| 1553 |
+
"epoch": 3.507936507936508,
|
| 1554 |
+
"grad_norm": 0.5784669157284802,
|
| 1555 |
+
"learning_rate": 2.532114095079137e-06,
|
| 1556 |
+
"loss": 0.5754,
|
| 1557 |
+
"step": 221
|
| 1558 |
+
},
|
| 1559 |
+
{
|
| 1560 |
+
"epoch": 3.5238095238095237,
|
| 1561 |
+
"grad_norm": 0.451433112279235,
|
| 1562 |
+
"learning_rate": 2.4839941896534027e-06,
|
| 1563 |
+
"loss": 0.6267,
|
| 1564 |
+
"step": 222
|
| 1565 |
+
},
|
| 1566 |
+
{
|
| 1567 |
+
"epoch": 3.5396825396825395,
|
| 1568 |
+
"grad_norm": 0.43076481074733924,
|
| 1569 |
+
"learning_rate": 2.4361843358828287e-06,
|
| 1570 |
+
"loss": 0.5873,
|
| 1571 |
+
"step": 223
|
| 1572 |
+
},
|
| 1573 |
+
{
|
| 1574 |
+
"epoch": 3.5555555555555554,
|
| 1575 |
+
"grad_norm": 0.5555243876307536,
|
| 1576 |
+
"learning_rate": 2.388690425456629e-06,
|
| 1577 |
+
"loss": 0.7168,
|
| 1578 |
+
"step": 224
|
| 1579 |
+
},
|
| 1580 |
+
{
|
| 1581 |
+
"epoch": 3.571428571428571,
|
| 1582 |
+
"grad_norm": 0.5140731956812183,
|
| 1583 |
+
"learning_rate": 2.341518311129781e-06,
|
| 1584 |
+
"loss": 0.5671,
|
| 1585 |
+
"step": 225
|
| 1586 |
+
},
|
| 1587 |
+
{
|
| 1588 |
+
"epoch": 3.5873015873015874,
|
| 1589 |
+
"grad_norm": 0.4829910245643812,
|
| 1590 |
+
"learning_rate": 2.2946738060017947e-06,
|
| 1591 |
+
"loss": 0.708,
|
| 1592 |
+
"step": 226
|
| 1593 |
+
},
|
| 1594 |
+
{
|
| 1595 |
+
"epoch": 3.6031746031746033,
|
| 1596 |
+
"grad_norm": 0.4822066553569954,
|
| 1597 |
+
"learning_rate": 2.24816268280033e-06,
|
| 1598 |
+
"loss": 0.653,
|
| 1599 |
+
"step": 227
|
| 1600 |
+
},
|
| 1601 |
+
{
|
| 1602 |
+
"epoch": 3.619047619047619,
|
| 1603 |
+
"grad_norm": 0.47281479364796025,
|
| 1604 |
+
"learning_rate": 2.2019906731698337e-06,
|
| 1605 |
+
"loss": 0.6494,
|
| 1606 |
+
"step": 228
|
| 1607 |
+
},
|
| 1608 |
+
{
|
| 1609 |
+
"epoch": 3.634920634920635,
|
| 1610 |
+
"grad_norm": 0.40568931340881786,
|
| 1611 |
+
"learning_rate": 2.156163466965218e-06,
|
| 1612 |
+
"loss": 0.6293,
|
| 1613 |
+
"step": 229
|
| 1614 |
+
},
|
| 1615 |
+
{
|
| 1616 |
+
"epoch": 3.6507936507936507,
|
| 1617 |
+
"grad_norm": 0.8300833772536724,
|
| 1618 |
+
"learning_rate": 2.110686711550678e-06,
|
| 1619 |
+
"loss": 0.7077,
|
| 1620 |
+
"step": 230
|
| 1621 |
+
},
|
| 1622 |
+
{
|
| 1623 |
+
"epoch": 3.6666666666666665,
|
| 1624 |
+
"grad_norm": 0.4061549756977283,
|
| 1625 |
+
"learning_rate": 2.0655660111037685e-06,
|
| 1626 |
+
"loss": 0.6339,
|
| 1627 |
+
"step": 231
|
| 1628 |
+
},
|
| 1629 |
+
{
|
| 1630 |
+
"epoch": 3.682539682539683,
|
| 1631 |
+
"grad_norm": 0.4255055033708699,
|
| 1632 |
+
"learning_rate": 2.0208069259247866e-06,
|
| 1633 |
+
"loss": 0.6577,
|
| 1634 |
+
"step": 232
|
| 1635 |
+
},
|
| 1636 |
+
{
|
| 1637 |
+
"epoch": 3.6984126984126986,
|
| 1638 |
+
"grad_norm": 0.43647553702799224,
|
| 1639 |
+
"learning_rate": 1.976414971751568e-06,
|
| 1640 |
+
"loss": 0.6029,
|
| 1641 |
+
"step": 233
|
| 1642 |
+
},
|
| 1643 |
+
{
|
| 1644 |
+
"epoch": 3.7142857142857144,
|
| 1645 |
+
"grad_norm": 0.4631401287763674,
|
| 1646 |
+
"learning_rate": 1.932395619079771e-06,
|
| 1647 |
+
"loss": 0.6777,
|
| 1648 |
+
"step": 234
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 3.7301587301587302,
|
| 1652 |
+
"grad_norm": 0.4306497798982157,
|
| 1653 |
+
"learning_rate": 1.8887542924887486e-06,
|
| 1654 |
+
"loss": 0.6996,
|
| 1655 |
+
"step": 235
|
| 1656 |
+
},
|
| 1657 |
+
{
|
| 1658 |
+
"epoch": 3.746031746031746,
|
| 1659 |
+
"grad_norm": 0.42608290640248714,
|
| 1660 |
+
"learning_rate": 1.8454963699730471e-06,
|
| 1661 |
+
"loss": 0.6727,
|
| 1662 |
+
"step": 236
|
| 1663 |
+
},
|
| 1664 |
+
{
|
| 1665 |
+
"epoch": 3.761904761904762,
|
| 1666 |
+
"grad_norm": 0.43950627489492866,
|
| 1667 |
+
"learning_rate": 1.802627182279687e-06,
|
| 1668 |
+
"loss": 0.5927,
|
| 1669 |
+
"step": 237
|
| 1670 |
+
},
|
| 1671 |
+
{
|
| 1672 |
+
"epoch": 3.7777777777777777,
|
| 1673 |
+
"grad_norm": 0.8150329270991283,
|
| 1674 |
+
"learning_rate": 1.760152012251241e-06,
|
| 1675 |
+
"loss": 0.5039,
|
| 1676 |
+
"step": 238
|
| 1677 |
+
},
|
| 1678 |
+
{
|
| 1679 |
+
"epoch": 3.7936507936507935,
|
| 1680 |
+
"grad_norm": 0.5428442791120296,
|
| 1681 |
+
"learning_rate": 1.7180760941748132e-06,
|
| 1682 |
+
"loss": 0.6682,
|
| 1683 |
+
"step": 239
|
| 1684 |
+
},
|
| 1685 |
+
{
|
| 1686 |
+
"epoch": 3.8095238095238093,
|
| 1687 |
+
"grad_norm": 0.7967588807434276,
|
| 1688 |
+
"learning_rate": 1.6764046131370142e-06,
|
| 1689 |
+
"loss": 0.7302,
|
| 1690 |
+
"step": 240
|
| 1691 |
+
},
|
| 1692 |
+
{
|
| 1693 |
+
"epoch": 3.825396825396825,
|
| 1694 |
+
"grad_norm": 0.4533126401638171,
|
| 1695 |
+
"learning_rate": 1.6351427043849955e-06,
|
| 1696 |
+
"loss": 0.6953,
|
| 1697 |
+
"step": 241
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"epoch": 3.8412698412698414,
|
| 1701 |
+
"grad_norm": 0.5989253712710739,
|
| 1702 |
+
"learning_rate": 1.5942954526936217e-06,
|
| 1703 |
+
"loss": 0.654,
|
| 1704 |
+
"step": 242
|
| 1705 |
+
},
|
| 1706 |
+
{
|
| 1707 |
+
"epoch": 3.857142857142857,
|
| 1708 |
+
"grad_norm": 0.5693019253502395,
|
| 1709 |
+
"learning_rate": 1.5538678917388638e-06,
|
| 1710 |
+
"loss": 0.6618,
|
| 1711 |
+
"step": 243
|
| 1712 |
+
},
|
| 1713 |
+
{
|
| 1714 |
+
"epoch": 3.873015873015873,
|
| 1715 |
+
"grad_norm": 0.47402524421174175,
|
| 1716 |
+
"learning_rate": 1.5138650034775004e-06,
|
| 1717 |
+
"loss": 0.5709,
|
| 1718 |
+
"step": 244
|
| 1719 |
+
},
|
| 1720 |
+
{
|
| 1721 |
+
"epoch": 3.888888888888889,
|
| 1722 |
+
"grad_norm": 0.4074162350507339,
|
| 1723 |
+
"learning_rate": 1.4742917175331644e-06,
|
| 1724 |
+
"loss": 0.7154,
|
| 1725 |
+
"step": 245
|
| 1726 |
+
},
|
| 1727 |
+
{
|
| 1728 |
+
"epoch": 3.9047619047619047,
|
| 1729 |
+
"grad_norm": 0.4421839212167129,
|
| 1730 |
+
"learning_rate": 1.4351529105888735e-06,
|
| 1731 |
+
"loss": 0.7655,
|
| 1732 |
+
"step": 246
|
| 1733 |
+
},
|
| 1734 |
+
{
|
| 1735 |
+
"epoch": 3.9206349206349205,
|
| 1736 |
+
"grad_norm": 0.45643272842877347,
|
| 1737 |
+
"learning_rate": 1.3964534057860652e-06,
|
| 1738 |
+
"loss": 0.5952,
|
| 1739 |
+
"step": 247
|
| 1740 |
+
},
|
| 1741 |
+
{
|
| 1742 |
+
"epoch": 3.9365079365079367,
|
| 1743 |
+
"grad_norm": 0.4769259453601474,
|
| 1744 |
+
"learning_rate": 1.3581979721302286e-06,
|
| 1745 |
+
"loss": 0.6536,
|
| 1746 |
+
"step": 248
|
| 1747 |
+
},
|
| 1748 |
+
{
|
| 1749 |
+
"epoch": 3.9523809523809526,
|
| 1750 |
+
"grad_norm": 0.5748113648603245,
|
| 1751 |
+
"learning_rate": 1.3203913239032074e-06,
|
| 1752 |
+
"loss": 0.6675,
|
| 1753 |
+
"step": 249
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 3.9682539682539684,
|
| 1757 |
+
"grad_norm": 0.5055584872741437,
|
| 1758 |
+
"learning_rate": 1.283038120082268e-06,
|
| 1759 |
+
"loss": 0.6196,
|
| 1760 |
+
"step": 250
|
| 1761 |
+
},
|
| 1762 |
+
{
|
| 1763 |
+
"epoch": 3.984126984126984,
|
| 1764 |
+
"grad_norm": 0.4492562647446275,
|
| 1765 |
+
"learning_rate": 1.2461429637659466e-06,
|
| 1766 |
+
"loss": 0.6194,
|
| 1767 |
+
"step": 251
|
| 1768 |
+
},
|
| 1769 |
+
{
|
| 1770 |
+
"epoch": 4.0,
|
| 1771 |
+
"grad_norm": 0.5069425974799285,
|
| 1772 |
+
"learning_rate": 1.2097104016068146e-06,
|
| 1773 |
+
"loss": 0.6338,
|
| 1774 |
+
"step": 252
|
| 1775 |
+
},
|
| 1776 |
+
{
|
| 1777 |
+
"epoch": 4.015873015873016,
|
| 1778 |
+
"grad_norm": 0.7325130340591377,
|
| 1779 |
+
"learning_rate": 1.1737449232511799e-06,
|
| 1780 |
+
"loss": 0.6351,
|
| 1781 |
+
"step": 253
|
| 1782 |
+
},
|
| 1783 |
+
{
|
| 1784 |
+
"epoch": 4.031746031746032,
|
| 1785 |
+
"grad_norm": 0.47508153140792514,
|
| 1786 |
+
"learning_rate": 1.1382509607858233e-06,
|
| 1787 |
+
"loss": 0.6771,
|
| 1788 |
+
"step": 254
|
| 1789 |
+
},
|
| 1790 |
+
{
|
| 1791 |
+
"epoch": 4.0476190476190474,
|
| 1792 |
+
"grad_norm": 0.42235140407913196,
|
| 1793 |
+
"learning_rate": 1.1032328881918237e-06,
|
| 1794 |
+
"loss": 0.6626,
|
| 1795 |
+
"step": 255
|
| 1796 |
+
},
|
| 1797 |
+
{
|
| 1798 |
+
"epoch": 4.063492063492063,
|
| 1799 |
+
"grad_norm": 0.5366768605947194,
|
| 1800 |
+
"learning_rate": 1.0686950208055486e-06,
|
| 1801 |
+
"loss": 0.6944,
|
| 1802 |
+
"step": 256
|
| 1803 |
+
},
|
| 1804 |
+
{
|
| 1805 |
+
"epoch": 4.079365079365079,
|
| 1806 |
+
"grad_norm": 0.48961048311306354,
|
| 1807 |
+
"learning_rate": 1.034641614786862e-06,
|
| 1808 |
+
"loss": 0.6248,
|
| 1809 |
+
"step": 257
|
| 1810 |
+
},
|
| 1811 |
+
{
|
| 1812 |
+
"epoch": 4.095238095238095,
|
| 1813 |
+
"grad_norm": 0.6474817812230599,
|
| 1814 |
+
"learning_rate": 1.0010768665946309e-06,
|
| 1815 |
+
"loss": 0.6044,
|
| 1816 |
+
"step": 258
|
| 1817 |
+
},
|
| 1818 |
+
{
|
| 1819 |
+
"epoch": 4.111111111111111,
|
| 1820 |
+
"grad_norm": 0.5770682040417152,
|
| 1821 |
+
"learning_rate": 9.680049124695973e-07,
|
| 1822 |
+
"loss": 0.6343,
|
| 1823 |
+
"step": 259
|
| 1824 |
+
},
|
| 1825 |
+
{
|
| 1826 |
+
"epoch": 4.1269841269841265,
|
| 1827 |
+
"grad_norm": 0.5366437759340096,
|
| 1828 |
+
"learning_rate": 9.35429827924652e-07,
|
| 1829 |
+
"loss": 0.6539,
|
| 1830 |
+
"step": 260
|
| 1831 |
+
},
|
| 1832 |
+
{
|
| 1833 |
+
"epoch": 4.142857142857143,
|
| 1834 |
+
"grad_norm": 0.6413478244014154,
|
| 1835 |
+
"learning_rate": 9.033556272426075e-07,
|
| 1836 |
+
"loss": 0.5768,
|
| 1837 |
+
"step": 261
|
| 1838 |
+
},
|
| 1839 |
+
{
|
| 1840 |
+
"epoch": 4.158730158730159,
|
| 1841 |
+
"grad_norm": 0.6609629042151937,
|
| 1842 |
+
"learning_rate": 8.717862629815099e-07,
|
| 1843 |
+
"loss": 0.66,
|
| 1844 |
+
"step": 262
|
| 1845 |
+
},
|
| 1846 |
+
{
|
| 1847 |
+
"epoch": 4.174603174603175,
|
| 1848 |
+
"grad_norm": 0.4842145086290818,
|
| 1849 |
+
"learning_rate": 8.407256254875573e-07,
|
| 1850 |
+
"loss": 0.5534,
|
| 1851 |
+
"step": 263
|
| 1852 |
+
},
|
| 1853 |
+
{
|
| 1854 |
+
"epoch": 4.190476190476191,
|
| 1855 |
+
"grad_norm": 0.42750956471439316,
|
| 1856 |
+
"learning_rate": 8.101775424156888e-07,
|
| 1857 |
+
"loss": 0.7396,
|
| 1858 |
+
"step": 264
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 4.2063492063492065,
|
| 1862 |
+
"grad_norm": 0.43019081247965996,
|
| 1863 |
+
"learning_rate": 7.801457782578947e-07,
|
| 1864 |
+
"loss": 0.574,
|
| 1865 |
+
"step": 265
|
| 1866 |
+
},
|
| 1867 |
+
{
|
| 1868 |
+
"epoch": 4.222222222222222,
|
| 1869 |
+
"grad_norm": 0.44008885788118995,
|
| 1870 |
+
"learning_rate": 7.506340338793111e-07,
|
| 1871 |
+
"loss": 0.6995,
|
| 1872 |
+
"step": 266
|
| 1873 |
+
},
|
| 1874 |
+
{
|
| 1875 |
+
"epoch": 4.238095238095238,
|
| 1876 |
+
"grad_norm": 0.46207221181078584,
|
| 1877 |
+
"learning_rate": 7.216459460621528e-07,
|
| 1878 |
+
"loss": 0.5681,
|
| 1879 |
+
"step": 267
|
| 1880 |
+
},
|
| 1881 |
+
{
|
| 1882 |
+
"epoch": 4.253968253968254,
|
| 1883 |
+
"grad_norm": 0.5615985950866916,
|
| 1884 |
+
"learning_rate": 6.931850870575563e-07,
|
| 1885 |
+
"loss": 0.6041,
|
| 1886 |
+
"step": 268
|
| 1887 |
+
},
|
| 1888 |
+
{
|
| 1889 |
+
"epoch": 4.26984126984127,
|
| 1890 |
+
"grad_norm": 0.7306441290684199,
|
| 1891 |
+
"learning_rate": 6.652549641453543e-07,
|
| 1892 |
+
"loss": 0.652,
|
| 1893 |
+
"step": 269
|
| 1894 |
+
},
|
| 1895 |
+
{
|
| 1896 |
+
"epoch": 4.285714285714286,
|
| 1897 |
+
"grad_norm": 0.6570180830204069,
|
| 1898 |
+
"learning_rate": 6.378590192018752e-07,
|
| 1899 |
+
"loss": 0.5264,
|
| 1900 |
+
"step": 270
|
| 1901 |
+
},
|
| 1902 |
+
{
|
| 1903 |
+
"epoch": 4.301587301587301,
|
| 1904 |
+
"grad_norm": 0.5038100342096954,
|
| 1905 |
+
"learning_rate": 6.110006282757897e-07,
|
| 1906 |
+
"loss": 0.6315,
|
| 1907 |
+
"step": 271
|
| 1908 |
+
},
|
| 1909 |
+
{
|
| 1910 |
+
"epoch": 4.317460317460317,
|
| 1911 |
+
"grad_norm": 0.49329707190738137,
|
| 1912 |
+
"learning_rate": 5.846831011720789e-07,
|
| 1913 |
+
"loss": 0.5652,
|
| 1914 |
+
"step": 272
|
| 1915 |
+
},
|
| 1916 |
+
{
|
| 1917 |
+
"epoch": 4.333333333333333,
|
| 1918 |
+
"grad_norm": 0.47592088200670746,
|
| 1919 |
+
"learning_rate": 5.589096810441574e-07,
|
| 1920 |
+
"loss": 0.5355,
|
| 1921 |
+
"step": 273
|
| 1922 |
+
},
|
| 1923 |
+
{
|
| 1924 |
+
"epoch": 4.349206349206349,
|
| 1925 |
+
"grad_norm": 0.41595101638818205,
|
| 1926 |
+
"learning_rate": 5.3368354399422e-07,
|
| 1927 |
+
"loss": 0.6575,
|
| 1928 |
+
"step": 274
|
| 1929 |
+
},
|
| 1930 |
+
{
|
| 1931 |
+
"epoch": 4.365079365079365,
|
| 1932 |
+
"grad_norm": 0.4900815198733562,
|
| 1933 |
+
"learning_rate": 5.090077986818365e-07,
|
| 1934 |
+
"loss": 0.5854,
|
| 1935 |
+
"step": 275
|
| 1936 |
+
},
|
| 1937 |
+
{
|
| 1938 |
+
"epoch": 4.380952380952381,
|
| 1939 |
+
"grad_norm": 0.5644693062570447,
|
| 1940 |
+
"learning_rate": 4.848854859408731e-07,
|
| 1941 |
+
"loss": 0.6625,
|
| 1942 |
+
"step": 276
|
| 1943 |
+
},
|
| 1944 |
+
{
|
| 1945 |
+
"epoch": 4.396825396825397,
|
| 1946 |
+
"grad_norm": 0.47377982529710666,
|
| 1947 |
+
"learning_rate": 4.613195784047653e-07,
|
| 1948 |
+
"loss": 0.596,
|
| 1949 |
+
"step": 277
|
| 1950 |
+
},
|
| 1951 |
+
{
|
| 1952 |
+
"epoch": 4.412698412698413,
|
| 1953 |
+
"grad_norm": 0.4584447745975199,
|
| 1954 |
+
"learning_rate": 4.3831298014019144e-07,
|
| 1955 |
+
"loss": 0.6383,
|
| 1956 |
+
"step": 278
|
| 1957 |
+
},
|
| 1958 |
+
{
|
| 1959 |
+
"epoch": 4.428571428571429,
|
| 1960 |
+
"grad_norm": 0.4565322356771251,
|
| 1961 |
+
"learning_rate": 4.1586852628920095e-07,
|
| 1962 |
+
"loss": 0.5529,
|
| 1963 |
+
"step": 279
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 4.444444444444445,
|
| 1967 |
+
"grad_norm": 0.47404711132247973,
|
| 1968 |
+
"learning_rate": 3.939889827198362e-07,
|
| 1969 |
+
"loss": 0.4942,
|
| 1970 |
+
"step": 280
|
| 1971 |
+
},
|
| 1972 |
+
{
|
| 1973 |
+
"epoch": 4.4603174603174605,
|
| 1974 |
+
"grad_norm": 0.45271106934439176,
|
| 1975 |
+
"learning_rate": 3.7267704568529015e-07,
|
| 1976 |
+
"loss": 0.5257,
|
| 1977 |
+
"step": 281
|
| 1978 |
+
},
|
| 1979 |
+
{
|
| 1980 |
+
"epoch": 4.476190476190476,
|
| 1981 |
+
"grad_norm": 0.6723472247553434,
|
| 1982 |
+
"learning_rate": 3.519353414916404e-07,
|
| 1983 |
+
"loss": 0.6823,
|
| 1984 |
+
"step": 282
|
| 1985 |
+
},
|
| 1986 |
+
{
|
| 1987 |
+
"epoch": 4.492063492063492,
|
| 1988 |
+
"grad_norm": 0.4048181921028701,
|
| 1989 |
+
"learning_rate": 3.3176642617420817e-07,
|
| 1990 |
+
"loss": 0.6414,
|
| 1991 |
+
"step": 283
|
| 1992 |
+
},
|
| 1993 |
+
{
|
| 1994 |
+
"epoch": 4.507936507936508,
|
| 1995 |
+
"grad_norm": 0.505253004014245,
|
| 1996 |
+
"learning_rate": 3.1217278518256844e-07,
|
| 1997 |
+
"loss": 0.6771,
|
| 1998 |
+
"step": 284
|
| 1999 |
+
},
|
| 2000 |
+
{
|
| 2001 |
+
"epoch": 4.523809523809524,
|
| 2002 |
+
"grad_norm": 0.43581438187229343,
|
| 2003 |
+
"learning_rate": 2.93156833074269e-07,
|
| 2004 |
+
"loss": 0.7079,
|
| 2005 |
+
"step": 285
|
| 2006 |
+
},
|
| 2007 |
+
{
|
| 2008 |
+
"epoch": 4.5396825396825395,
|
| 2009 |
+
"grad_norm": 0.4401525579247126,
|
| 2010 |
+
"learning_rate": 2.7472091321728067e-07,
|
| 2011 |
+
"loss": 0.5173,
|
| 2012 |
+
"step": 286
|
| 2013 |
+
},
|
| 2014 |
+
{
|
| 2015 |
+
"epoch": 4.555555555555555,
|
| 2016 |
+
"grad_norm": 0.5337113220659214,
|
| 2017 |
+
"learning_rate": 2.568672975012154e-07,
|
| 2018 |
+
"loss": 0.5528,
|
| 2019 |
+
"step": 287
|
| 2020 |
+
},
|
| 2021 |
+
{
|
| 2022 |
+
"epoch": 4.571428571428571,
|
| 2023 |
+
"grad_norm": 0.49778493285427783,
|
| 2024 |
+
"learning_rate": 2.3959818605736095e-07,
|
| 2025 |
+
"loss": 0.5663,
|
| 2026 |
+
"step": 288
|
| 2027 |
+
},
|
| 2028 |
+
{
|
| 2029 |
+
"epoch": 4.587301587301587,
|
| 2030 |
+
"grad_norm": 0.40499062553885884,
|
| 2031 |
+
"learning_rate": 2.229157069875537e-07,
|
| 2032 |
+
"loss": 0.6321,
|
| 2033 |
+
"step": 289
|
| 2034 |
+
},
|
| 2035 |
+
{
|
| 2036 |
+
"epoch": 4.603174603174603,
|
| 2037 |
+
"grad_norm": 0.4127656173064128,
|
| 2038 |
+
"learning_rate": 2.068219161019297e-07,
|
| 2039 |
+
"loss": 0.4887,
|
| 2040 |
+
"step": 290
|
| 2041 |
+
},
|
| 2042 |
+
{
|
| 2043 |
+
"epoch": 4.619047619047619,
|
| 2044 |
+
"grad_norm": 0.5639846795868625,
|
| 2045 |
+
"learning_rate": 1.9131879666558385e-07,
|
| 2046 |
+
"loss": 0.6286,
|
| 2047 |
+
"step": 291
|
| 2048 |
+
},
|
| 2049 |
+
{
|
| 2050 |
+
"epoch": 4.634920634920634,
|
| 2051 |
+
"grad_norm": 0.5278765172669099,
|
| 2052 |
+
"learning_rate": 1.7640825915416994e-07,
|
| 2053 |
+
"loss": 0.5375,
|
| 2054 |
+
"step": 292
|
| 2055 |
+
},
|
| 2056 |
+
{
|
| 2057 |
+
"epoch": 4.650793650793651,
|
| 2058 |
+
"grad_norm": 0.39500499979498543,
|
| 2059 |
+
"learning_rate": 1.6209214101846394e-07,
|
| 2060 |
+
"loss": 0.5712,
|
| 2061 |
+
"step": 293
|
| 2062 |
+
},
|
| 2063 |
+
{
|
| 2064 |
+
"epoch": 4.666666666666667,
|
| 2065 |
+
"grad_norm": 0.4763017000664341,
|
| 2066 |
+
"learning_rate": 1.4837220645793905e-07,
|
| 2067 |
+
"loss": 0.6848,
|
| 2068 |
+
"step": 294
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 4.682539682539683,
|
| 2072 |
+
"grad_norm": 0.44371008674048357,
|
| 2073 |
+
"learning_rate": 1.3525014620335786e-07,
|
| 2074 |
+
"loss": 0.671,
|
| 2075 |
+
"step": 295
|
| 2076 |
+
},
|
| 2077 |
+
{
|
| 2078 |
+
"epoch": 4.698412698412699,
|
| 2079 |
+
"grad_norm": 0.48437863773460277,
|
| 2080 |
+
"learning_rate": 1.2272757730841744e-07,
|
| 2081 |
+
"loss": 0.7189,
|
| 2082 |
+
"step": 296
|
| 2083 |
+
},
|
| 2084 |
+
{
|
| 2085 |
+
"epoch": 4.714285714285714,
|
| 2086 |
+
"grad_norm": 0.466188982528453,
|
| 2087 |
+
"learning_rate": 1.1080604295048203e-07,
|
| 2088 |
+
"loss": 0.5655,
|
| 2089 |
+
"step": 297
|
| 2090 |
+
},
|
| 2091 |
+
{
|
| 2092 |
+
"epoch": 4.73015873015873,
|
| 2093 |
+
"grad_norm": 0.5243644807437886,
|
| 2094 |
+
"learning_rate": 9.948701224041124e-08,
|
| 2095 |
+
"loss": 0.678,
|
| 2096 |
+
"step": 298
|
| 2097 |
+
},
|
| 2098 |
+
{
|
| 2099 |
+
"epoch": 4.746031746031746,
|
| 2100 |
+
"grad_norm": 0.41383827022733516,
|
| 2101 |
+
"learning_rate": 8.877188004152104e-08,
|
| 2102 |
+
"loss": 0.677,
|
| 2103 |
+
"step": 299
|
| 2104 |
+
},
|
| 2105 |
+
{
|
| 2106 |
+
"epoch": 4.761904761904762,
|
| 2107 |
+
"grad_norm": 0.4845407723856424,
|
| 2108 |
+
"learning_rate": 7.866196679768956e-08,
|
| 2109 |
+
"loss": 0.5844,
|
| 2110 |
+
"step": 300
|
| 2111 |
+
},
|
| 2112 |
+
{
|
| 2113 |
+
"epoch": 4.777777777777778,
|
| 2114 |
+
"grad_norm": 0.4125199836030617,
|
| 2115 |
+
"learning_rate": 6.91585183706428e-08,
|
| 2116 |
+
"loss": 0.5936,
|
| 2117 |
+
"step": 301
|
| 2118 |
+
},
|
| 2119 |
+
{
|
| 2120 |
+
"epoch": 4.7936507936507935,
|
| 2121 |
+
"grad_norm": 0.42044450878999223,
|
| 2122 |
+
"learning_rate": 6.02627058864158e-08,
|
| 2123 |
+
"loss": 0.6322,
|
| 2124 |
+
"step": 302
|
| 2125 |
+
},
|
| 2126 |
+
{
|
| 2127 |
+
"epoch": 4.809523809523809,
|
| 2128 |
+
"grad_norm": 0.43796427028494905,
|
| 2129 |
+
"learning_rate": 5.19756255910403e-08,
|
| 2130 |
+
"loss": 0.6218,
|
| 2131 |
+
"step": 303
|
| 2132 |
+
},
|
| 2133 |
+
{
|
| 2134 |
+
"epoch": 4.825396825396825,
|
| 2135 |
+
"grad_norm": 0.4943567439631721,
|
| 2136 |
+
"learning_rate": 4.429829871545055e-08,
|
| 2137 |
+
"loss": 0.6154,
|
| 2138 |
+
"step": 304
|
| 2139 |
+
},
|
| 2140 |
+
{
|
| 2141 |
+
"epoch": 4.841269841269841,
|
| 2142 |
+
"grad_norm": 0.4041231594209634,
|
| 2143 |
+
"learning_rate": 3.7231671349634015e-08,
|
| 2144 |
+
"loss": 0.5363,
|
| 2145 |
+
"step": 305
|
| 2146 |
+
},
|
| 2147 |
+
{
|
| 2148 |
+
"epoch": 4.857142857142857,
|
| 2149 |
+
"grad_norm": 0.4244498098386566,
|
| 2150 |
+
"learning_rate": 3.077661432604184e-08,
|
| 2151 |
+
"loss": 0.5705,
|
| 2152 |
+
"step": 306
|
| 2153 |
+
},
|
| 2154 |
+
{
|
| 2155 |
+
"epoch": 4.8730158730158735,
|
| 2156 |
+
"grad_norm": 0.4028466383216641,
|
| 2157 |
+
"learning_rate": 2.4933923112279712e-08,
|
| 2158 |
+
"loss": 0.6748,
|
| 2159 |
+
"step": 307
|
| 2160 |
+
},
|
| 2161 |
+
{
|
| 2162 |
+
"epoch": 4.888888888888889,
|
| 2163 |
+
"grad_norm": 0.4337309649014186,
|
| 2164 |
+
"learning_rate": 1.9704317713076236e-08,
|
| 2165 |
+
"loss": 0.7078,
|
| 2166 |
+
"step": 308
|
| 2167 |
+
},
|
| 2168 |
+
{
|
| 2169 |
+
"epoch": 4.904761904761905,
|
| 2170 |
+
"grad_norm": 0.4603602645522074,
|
| 2171 |
+
"learning_rate": 1.508844258155728e-08,
|
| 2172 |
+
"loss": 0.6406,
|
| 2173 |
+
"step": 309
|
| 2174 |
+
},
|
| 2175 |
+
{
|
| 2176 |
+
"epoch": 4.920634920634921,
|
| 2177 |
+
"grad_norm": 0.408641845898798,
|
| 2178 |
+
"learning_rate": 1.1086866539830044e-08,
|
| 2179 |
+
"loss": 0.6818,
|
| 2180 |
+
"step": 310
|
| 2181 |
+
},
|
| 2182 |
+
{
|
| 2183 |
+
"epoch": 4.936507936507937,
|
| 2184 |
+
"grad_norm": 0.4380910149550681,
|
| 2185 |
+
"learning_rate": 7.700082708883006e-09,
|
| 2186 |
+
"loss": 0.6386,
|
| 2187 |
+
"step": 311
|
| 2188 |
+
},
|
| 2189 |
+
{
|
| 2190 |
+
"epoch": 4.9523809523809526,
|
| 2191 |
+
"grad_norm": 0.4375524382908019,
|
| 2192 |
+
"learning_rate": 4.928508447821223e-09,
|
| 2193 |
+
"loss": 0.6198,
|
| 2194 |
+
"step": 312
|
| 2195 |
+
},
|
| 2196 |
+
{
|
| 2197 |
+
"epoch": 4.968253968253968,
|
| 2198 |
+
"grad_norm": 0.4660304819447141,
|
| 2199 |
+
"learning_rate": 2.7724853024324594e-09,
|
| 2200 |
+
"loss": 0.5728,
|
| 2201 |
+
"step": 313
|
| 2202 |
+
},
|
| 2203 |
+
{
|
| 2204 |
+
"epoch": 4.984126984126984,
|
| 2205 |
+
"grad_norm": 0.5605867904986561,
|
| 2206 |
+
"learning_rate": 1.2322789630997422e-09,
|
| 2207 |
+
"loss": 0.6566,
|
| 2208 |
+
"step": 314
|
| 2209 |
+
},
|
| 2210 |
+
{
|
| 2211 |
+
"epoch": 5.0,
|
| 2212 |
+
"grad_norm": 0.47483560805428604,
|
| 2213 |
+
"learning_rate": 3.080792320564463e-10,
|
| 2214 |
+
"loss": 0.5612,
|
| 2215 |
+
"step": 315
|
| 2216 |
+
},
|
| 2217 |
+
{
|
| 2218 |
+
"epoch": 5.0,
|
| 2219 |
+
"step": 315,
|
| 2220 |
+
"total_flos": 78480301031424.0,
|
| 2221 |
+
"train_loss": 0.7362074042123462,
|
| 2222 |
+
"train_runtime": 9541.1641,
|
| 2223 |
+
"train_samples_per_second": 0.524,
|
| 2224 |
+
"train_steps_per_second": 0.033
|
| 2225 |
+
}
|
| 2226 |
+
],
|
| 2227 |
+
"logging_steps": 1,
|
| 2228 |
+
"max_steps": 315,
|
| 2229 |
+
"num_input_tokens_seen": 0,
|
| 2230 |
+
"num_train_epochs": 5,
|
| 2231 |
+
"save_steps": 500,
|
| 2232 |
+
"stateful_callbacks": {
|
| 2233 |
+
"TrainerControl": {
|
| 2234 |
+
"args": {
|
| 2235 |
+
"should_epoch_stop": false,
|
| 2236 |
+
"should_evaluate": false,
|
| 2237 |
+
"should_log": false,
|
| 2238 |
+
"should_save": true,
|
| 2239 |
+
"should_training_stop": true
|
| 2240 |
+
},
|
| 2241 |
+
"attributes": {}
|
| 2242 |
+
}
|
| 2243 |
+
},
|
| 2244 |
+
"total_flos": 78480301031424.0,
|
| 2245 |
+
"train_batch_size": 2,
|
| 2246 |
+
"trial_name": null,
|
| 2247 |
+
"trial_params": null
|
| 2248 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f8697730d2011a093c7921daeffcc9edef31a875a7d40136d2139174d1ed35b8
|
| 3 |
+
size 7672
|
training_loss.png
ADDED
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|