End of training
Browse files- README.md +2 -2
- all_results.json +13 -13
- eval_layer_index_28_mat_mlp.c_proj_results.json +9 -0
- eval_layer_index_29_mat_mlp.c_proj_results.json +7 -7
- model-00001-of-00002.safetensors +2 -2
- model.safetensors.index.json +5 -13
- runs/Mar13_14-25-10_demo2/events.out.tfevents.1741903489.demo2.193067.4 +3 -0
- runs/Mar13_14-25-10_demo2/events.out.tfevents.1741903699.demo2.193067.5 +3 -0
- runs/Mar13_16-43-27_demo2/events.out.tfevents.1741910387.demo2.205134.0 +3 -0
- runs/Mar13_16-43-27_demo2/events.out.tfevents.1741922968.demo2.205134.1 +3 -0
- train_layer_index_28_mat_mlp.c_proj_results.json +8 -0
- train_layer_index_29_mat_mlp.c_proj_results.json +6 -6
- trainer_state.json +1188 -12
- training_args.bin +1 -1
README.md
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
---
|
| 2 |
base_model: LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct
|
| 3 |
-
datasets:
|
| 4 |
library_name: transformers
|
| 5 |
model_name: output_exaone_tucker_rank16_newdata_b1_g2_gradient2_e1_b512
|
| 6 |
tags:
|
|
@@ -12,7 +12,7 @@ licence: license
|
|
| 12 |
|
| 13 |
# Model Card for output_exaone_tucker_rank16_newdata_b1_g2_gradient2_e1_b512
|
| 14 |
|
| 15 |
-
This model is a fine-tuned version of [LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct](https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct) on the [
|
| 16 |
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 17 |
|
| 18 |
## Quick start
|
|
|
|
| 1 |
---
|
| 2 |
base_model: LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct
|
| 3 |
+
datasets: testing_layer_29_Matrix_mlp.c_proj
|
| 4 |
library_name: transformers
|
| 5 |
model_name: output_exaone_tucker_rank16_newdata_b1_g2_gradient2_e1_b512
|
| 6 |
tags:
|
|
|
|
| 12 |
|
| 13 |
# Model Card for output_exaone_tucker_rank16_newdata_b1_g2_gradient2_e1_b512
|
| 14 |
|
| 15 |
+
This model is a fine-tuned version of [LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct](https://huggingface.co/LGAI-EXAONE/EXAONE-3.5-2.4B-Instruct) on the [testing_layer_29_Matrix_mlp.c_proj](https://huggingface.co/datasets/testing_layer_29_Matrix_mlp.c_proj) dataset.
|
| 16 |
It has been trained using [TRL](https://github.com/huggingface/trl).
|
| 17 |
|
| 18 |
## Quick start
|
all_results.json
CHANGED
|
@@ -1,15 +1,15 @@
|
|
| 1 |
{
|
| 2 |
-
"eval_accuracy": 0.
|
| 3 |
-
"eval_loss":
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples":
|
| 6 |
-
"eval_samples_per_second": 28.
|
| 7 |
-
"eval_steps_per_second": 14.
|
| 8 |
-
"perplexity":
|
| 9 |
-
"total_flos":
|
| 10 |
-
"train_loss":
|
| 11 |
-
"train_runtime":
|
| 12 |
-
"train_samples":
|
| 13 |
-
"train_samples_per_second":
|
| 14 |
-
"train_steps_per_second":
|
| 15 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"eval_accuracy": 0.8085339415160145,
|
| 3 |
+
"eval_loss": 1.681640625,
|
| 4 |
+
"eval_runtime": 927.2318,
|
| 5 |
+
"eval_samples": 25971,
|
| 6 |
+
"eval_samples_per_second": 28.009,
|
| 7 |
+
"eval_steps_per_second": 14.005,
|
| 8 |
+
"perplexity": 5.374366061438169,
|
| 9 |
+
"total_flos": 9.3961326985583e+17,
|
| 10 |
+
"train_loss": 1.7179553695028122,
|
| 11 |
+
"train_runtime": 11634.1539,
|
| 12 |
+
"train_samples": 294393,
|
| 13 |
+
"train_samples_per_second": 25.304,
|
| 14 |
+
"train_steps_per_second": 6.326
|
| 15 |
}
|
eval_layer_index_28_mat_mlp.c_proj_results.json
ADDED
|
@@ -0,0 +1,9 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"eval_accuracy": 0.7446557896577809,
|
| 3 |
+
"eval_loss": 5.17578125,
|
| 4 |
+
"eval_runtime": 182.4709,
|
| 5 |
+
"eval_samples": 5103,
|
| 6 |
+
"eval_samples_per_second": 27.966,
|
| 7 |
+
"eval_steps_per_second": 13.986,
|
| 8 |
+
"perplexity": 176.93479059943397
|
| 9 |
+
}
|
eval_layer_index_29_mat_mlp.c_proj_results.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
-
"eval_accuracy": 0.
|
| 3 |
-
"eval_loss":
|
| 4 |
-
"eval_runtime":
|
| 5 |
-
"eval_samples":
|
| 6 |
-
"eval_samples_per_second": 28.
|
| 7 |
-
"eval_steps_per_second": 14.
|
| 8 |
-
"perplexity":
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"eval_accuracy": 0.8085339415160145,
|
| 3 |
+
"eval_loss": 1.681640625,
|
| 4 |
+
"eval_runtime": 927.2318,
|
| 5 |
+
"eval_samples": 25971,
|
| 6 |
+
"eval_samples_per_second": 28.009,
|
| 7 |
+
"eval_steps_per_second": 14.005,
|
| 8 |
+
"perplexity": 5.374366061438169
|
| 9 |
}
|
model-00001-of-00002.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:b8f8f30ae6f98975b62efe02993f798af7fee6c3dd2a00c9213c8e4750246109
|
| 3 |
+
size 4701521288
|
model.safetensors.index.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
-
"total_size":
|
| 4 |
},
|
| 5 |
"weight_map": {
|
| 6 |
"lm_head.weight": "model-00002-of-00002.safetensors",
|
|
@@ -202,18 +202,10 @@
|
|
| 202 |
"transformer.h.28.mlp.c_fc_0.weight": "model-00001-of-00002.safetensors",
|
| 203 |
"transformer.h.28.mlp.c_fc_1.weight": "model-00001-of-00002.safetensors",
|
| 204 |
"transformer.h.28.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
|
| 205 |
-
"transformer.h.29.attn.attention.k_proj.
|
| 206 |
-
"transformer.h.29.attn.attention.
|
| 207 |
-
"transformer.h.29.attn.attention.
|
| 208 |
-
"transformer.h.29.attn.attention.
|
| 209 |
-
"transformer.h.29.attn.attention.out_proj.tucker_factors.0": "model-00001-of-00002.safetensors",
|
| 210 |
-
"transformer.h.29.attn.attention.out_proj.tucker_factors.1": "model-00001-of-00002.safetensors",
|
| 211 |
-
"transformer.h.29.attn.attention.q_proj.tucker_core": "model-00001-of-00002.safetensors",
|
| 212 |
-
"transformer.h.29.attn.attention.q_proj.tucker_factors.0": "model-00001-of-00002.safetensors",
|
| 213 |
-
"transformer.h.29.attn.attention.q_proj.tucker_factors.1": "model-00001-of-00002.safetensors",
|
| 214 |
-
"transformer.h.29.attn.attention.v_proj.tucker_core": "model-00001-of-00002.safetensors",
|
| 215 |
-
"transformer.h.29.attn.attention.v_proj.tucker_factors.0": "model-00001-of-00002.safetensors",
|
| 216 |
-
"transformer.h.29.attn.attention.v_proj.tucker_factors.1": "model-00001-of-00002.safetensors",
|
| 217 |
"transformer.h.29.ln_1.weight": "model-00001-of-00002.safetensors",
|
| 218 |
"transformer.h.29.ln_2.weight": "model-00001-of-00002.safetensors",
|
| 219 |
"transformer.h.29.mlp.c_fc_0.tucker_core": "model-00001-of-00002.safetensors",
|
|
|
|
| 1 |
{
|
| 2 |
"metadata": {
|
| 3 |
+
"total_size": 5225777664
|
| 4 |
},
|
| 5 |
"weight_map": {
|
| 6 |
"lm_head.weight": "model-00002-of-00002.safetensors",
|
|
|
|
| 202 |
"transformer.h.28.mlp.c_fc_0.weight": "model-00001-of-00002.safetensors",
|
| 203 |
"transformer.h.28.mlp.c_fc_1.weight": "model-00001-of-00002.safetensors",
|
| 204 |
"transformer.h.28.mlp.c_proj.weight": "model-00001-of-00002.safetensors",
|
| 205 |
+
"transformer.h.29.attn.attention.k_proj.weight": "model-00001-of-00002.safetensors",
|
| 206 |
+
"transformer.h.29.attn.attention.out_proj.weight": "model-00001-of-00002.safetensors",
|
| 207 |
+
"transformer.h.29.attn.attention.q_proj.weight": "model-00001-of-00002.safetensors",
|
| 208 |
+
"transformer.h.29.attn.attention.v_proj.weight": "model-00001-of-00002.safetensors",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 209 |
"transformer.h.29.ln_1.weight": "model-00001-of-00002.safetensors",
|
| 210 |
"transformer.h.29.ln_2.weight": "model-00001-of-00002.safetensors",
|
| 211 |
"transformer.h.29.mlp.c_fc_0.tucker_core": "model-00001-of-00002.safetensors",
|
runs/Mar13_14-25-10_demo2/events.out.tfevents.1741903489.demo2.193067.4
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:98d60275642b11b67444ab2bc0c1d429ffdeb9f2a0f4c8c6a208ca24813230bb
|
| 3 |
+
size 6413
|
runs/Mar13_14-25-10_demo2/events.out.tfevents.1741903699.demo2.193067.5
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aa3219ec1a2cdfd00e35c7b9d6b6e05de2b2107fcce7ba2e0cfb7d11418503ff
|
| 3 |
+
size 467
|
runs/Mar13_16-43-27_demo2/events.out.tfevents.1741910387.demo2.205134.0
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:d7694a82a8f44650110657d9e9daee591aed308ebd41940f7957ef51a36802a9
|
| 3 |
+
size 47468
|
runs/Mar13_16-43-27_demo2/events.out.tfevents.1741922968.demo2.205134.1
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6dbeef7de6b74f1c6c10a1141bd28c7edcf66ccac977d5fd86516e67f984805e
|
| 3 |
+
size 481
|
train_layer_index_28_mat_mlp.c_proj_results.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"total_flos": 21926516883456.0,
|
| 3 |
+
"train_loss": 5.8623046875,
|
| 4 |
+
"train_runtime": 0.353,
|
| 5 |
+
"train_samples": 76256,
|
| 6 |
+
"train_samples_per_second": 22.664,
|
| 7 |
+
"train_steps_per_second": 5.666
|
| 8 |
+
}
|
train_layer_index_29_mat_mlp.c_proj_results.json
CHANGED
|
@@ -1,8 +1,8 @@
|
|
| 1 |
{
|
| 2 |
-
"total_flos":
|
| 3 |
-
"train_loss":
|
| 4 |
-
"train_runtime":
|
| 5 |
-
"train_samples":
|
| 6 |
-
"train_samples_per_second":
|
| 7 |
-
"train_steps_per_second":
|
| 8 |
}
|
|
|
|
| 1 |
{
|
| 2 |
+
"total_flos": 9.3961326985583e+17,
|
| 3 |
+
"train_loss": 1.7179553695028122,
|
| 4 |
+
"train_runtime": 11634.1539,
|
| 5 |
+
"train_samples": 294393,
|
| 6 |
+
"train_samples_per_second": 25.304,
|
| 7 |
+
"train_steps_per_second": 6.326
|
| 8 |
}
|
trainer_state.json
CHANGED
|
@@ -1,26 +1,1202 @@
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
-
"epoch": 0.
|
| 5 |
"eval_steps": 500,
|
| 6 |
-
"global_step":
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
-
"epoch": 0.
|
| 13 |
-
"
|
| 14 |
-
"
|
| 15 |
-
"
|
| 16 |
-
"
|
| 17 |
-
"
|
| 18 |
-
|
| 19 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 20 |
}
|
| 21 |
],
|
| 22 |
"logging_steps": 500,
|
| 23 |
-
"max_steps":
|
| 24 |
"num_input_tokens_seen": 0,
|
| 25 |
"num_train_epochs": 1,
|
| 26 |
"save_steps": 0,
|
|
@@ -36,7 +1212,7 @@
|
|
| 36 |
"attributes": {}
|
| 37 |
}
|
| 38 |
},
|
| 39 |
-
"total_flos":
|
| 40 |
"train_batch_size": 1,
|
| 41 |
"trial_name": null,
|
| 42 |
"trial_params": null
|
|
|
|
| 1 |
{
|
| 2 |
"best_metric": null,
|
| 3 |
"best_model_checkpoint": null,
|
| 4 |
+
"epoch": 0.9999932063832823,
|
| 5 |
"eval_steps": 500,
|
| 6 |
+
"global_step": 73598,
|
| 7 |
"is_hyper_param_search": false,
|
| 8 |
"is_local_process_zero": true,
|
| 9 |
"is_world_process_zero": true,
|
| 10 |
"log_history": [
|
| 11 |
{
|
| 12 |
+
"epoch": 0.006793616717732019,
|
| 13 |
+
"grad_norm": 2.0208959579467773,
|
| 14 |
+
"learning_rate": 0.0004966371368787195,
|
| 15 |
+
"loss": 2.0427,
|
| 16 |
+
"mean_token_accuracy": 0.5520137556195259,
|
| 17 |
+
"step": 500
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.013587233435464039,
|
| 21 |
+
"grad_norm": 1.8117527961730957,
|
| 22 |
+
"learning_rate": 0.0004932403054430827,
|
| 23 |
+
"loss": 1.8742,
|
| 24 |
+
"mean_token_accuracy": 0.5727833653092385,
|
| 25 |
+
"step": 1000
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.020380850153196056,
|
| 29 |
+
"grad_norm": 1.5713497400283813,
|
| 30 |
+
"learning_rate": 0.0004898434740074459,
|
| 31 |
+
"loss": 1.8367,
|
| 32 |
+
"mean_token_accuracy": 0.5783296891450882,
|
| 33 |
+
"step": 1500
|
| 34 |
+
},
|
| 35 |
+
{
|
| 36 |
+
"epoch": 0.027174466870928077,
|
| 37 |
+
"grad_norm": 1.6636950969696045,
|
| 38 |
+
"learning_rate": 0.00048644664257180905,
|
| 39 |
+
"loss": 1.8238,
|
| 40 |
+
"mean_token_accuracy": 0.5810909342765808,
|
| 41 |
+
"step": 2000
|
| 42 |
+
},
|
| 43 |
+
{
|
| 44 |
+
"epoch": 0.033968083588660096,
|
| 45 |
+
"grad_norm": 1.525341272354126,
|
| 46 |
+
"learning_rate": 0.0004830498111361722,
|
| 47 |
+
"loss": 1.8096,
|
| 48 |
+
"mean_token_accuracy": 0.5805237484574318,
|
| 49 |
+
"step": 2500
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.04076170030639211,
|
| 53 |
+
"grad_norm": 1.7869559526443481,
|
| 54 |
+
"learning_rate": 0.00047965297970053533,
|
| 55 |
+
"loss": 1.8145,
|
| 56 |
+
"mean_token_accuracy": 0.5809020410776138,
|
| 57 |
+
"step": 3000
|
| 58 |
+
},
|
| 59 |
+
{
|
| 60 |
+
"epoch": 0.04755531702412413,
|
| 61 |
+
"grad_norm": 1.389845848083496,
|
| 62 |
+
"learning_rate": 0.0004762561482648985,
|
| 63 |
+
"loss": 1.7937,
|
| 64 |
+
"mean_token_accuracy": 0.5843603150844574,
|
| 65 |
+
"step": 3500
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"epoch": 0.054348933741856155,
|
| 69 |
+
"grad_norm": 1.1970040798187256,
|
| 70 |
+
"learning_rate": 0.00047285931682926167,
|
| 71 |
+
"loss": 1.7926,
|
| 72 |
+
"mean_token_accuracy": 0.5820418976843357,
|
| 73 |
+
"step": 4000
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.06114255045958817,
|
| 77 |
+
"grad_norm": 1.4389044046401978,
|
| 78 |
+
"learning_rate": 0.00046946248539362484,
|
| 79 |
+
"loss": 1.7622,
|
| 80 |
+
"mean_token_accuracy": 0.5891672801971436,
|
| 81 |
+
"step": 4500
|
| 82 |
+
},
|
| 83 |
+
{
|
| 84 |
+
"epoch": 0.06793616717732019,
|
| 85 |
+
"grad_norm": 1.177388072013855,
|
| 86 |
+
"learning_rate": 0.000466065653957988,
|
| 87 |
+
"loss": 1.7738,
|
| 88 |
+
"mean_token_accuracy": 0.5879195163547992,
|
| 89 |
+
"step": 5000
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.07472978389505221,
|
| 93 |
+
"grad_norm": 1.3432203531265259,
|
| 94 |
+
"learning_rate": 0.00046268240984809374,
|
| 95 |
+
"loss": 1.7767,
|
| 96 |
+
"mean_token_accuracy": 0.5867721286118031,
|
| 97 |
+
"step": 5500
|
| 98 |
+
},
|
| 99 |
+
{
|
| 100 |
+
"epoch": 0.08152340061278422,
|
| 101 |
+
"grad_norm": 1.360933542251587,
|
| 102 |
+
"learning_rate": 0.0004592855784124569,
|
| 103 |
+
"loss": 1.7767,
|
| 104 |
+
"mean_token_accuracy": 0.5867527797520161,
|
| 105 |
+
"step": 6000
|
| 106 |
+
},
|
| 107 |
+
{
|
| 108 |
+
"epoch": 0.08831701733051625,
|
| 109 |
+
"grad_norm": 1.1908655166625977,
|
| 110 |
+
"learning_rate": 0.0004559023343025626,
|
| 111 |
+
"loss": 1.7606,
|
| 112 |
+
"mean_token_accuracy": 0.5915117806196213,
|
| 113 |
+
"step": 6500
|
| 114 |
+
},
|
| 115 |
+
{
|
| 116 |
+
"epoch": 0.09511063404824827,
|
| 117 |
+
"grad_norm": 1.1951971054077148,
|
| 118 |
+
"learning_rate": 0.00045250550286692574,
|
| 119 |
+
"loss": 1.7629,
|
| 120 |
+
"mean_token_accuracy": 0.5892836402058601,
|
| 121 |
+
"step": 7000
|
| 122 |
+
},
|
| 123 |
+
{
|
| 124 |
+
"epoch": 0.10190425076598028,
|
| 125 |
+
"grad_norm": 1.207234501838684,
|
| 126 |
+
"learning_rate": 0.00044912225875703147,
|
| 127 |
+
"loss": 1.7618,
|
| 128 |
+
"mean_token_accuracy": 0.5882115732729435,
|
| 129 |
+
"step": 7500
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.10869786748371231,
|
| 133 |
+
"grad_norm": 1.1731985807418823,
|
| 134 |
+
"learning_rate": 0.0004457254273213946,
|
| 135 |
+
"loss": 1.753,
|
| 136 |
+
"mean_token_accuracy": 0.5895786600708961,
|
| 137 |
+
"step": 8000
|
| 138 |
+
},
|
| 139 |
+
{
|
| 140 |
+
"epoch": 0.11549148420144432,
|
| 141 |
+
"grad_norm": 1.3629071712493896,
|
| 142 |
+
"learning_rate": 0.0004423421832115003,
|
| 143 |
+
"loss": 1.7627,
|
| 144 |
+
"mean_token_accuracy": 0.587433840662241,
|
| 145 |
+
"step": 8500
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.12228510091917634,
|
| 149 |
+
"grad_norm": 1.2118574380874634,
|
| 150 |
+
"learning_rate": 0.0004389453517758635,
|
| 151 |
+
"loss": 1.7529,
|
| 152 |
+
"mean_token_accuracy": 0.5890023435950279,
|
| 153 |
+
"step": 9000
|
| 154 |
+
},
|
| 155 |
+
{
|
| 156 |
+
"epoch": 0.12907871763690837,
|
| 157 |
+
"grad_norm": 1.1447389125823975,
|
| 158 |
+
"learning_rate": 0.0004355621076659692,
|
| 159 |
+
"loss": 1.7433,
|
| 160 |
+
"mean_token_accuracy": 0.5922272650897503,
|
| 161 |
+
"step": 9500
|
| 162 |
+
},
|
| 163 |
+
{
|
| 164 |
+
"epoch": 0.13587233435464038,
|
| 165 |
+
"grad_norm": 1.1409871578216553,
|
| 166 |
+
"learning_rate": 0.00043216527623033237,
|
| 167 |
+
"loss": 1.7479,
|
| 168 |
+
"mean_token_accuracy": 0.5903287023007869,
|
| 169 |
+
"step": 10000
|
| 170 |
+
},
|
| 171 |
+
{
|
| 172 |
+
"epoch": 0.1426659510723724,
|
| 173 |
+
"grad_norm": 1.2514444589614868,
|
| 174 |
+
"learning_rate": 0.00042878203212043804,
|
| 175 |
+
"loss": 1.7471,
|
| 176 |
+
"mean_token_accuracy": 0.5892516044676304,
|
| 177 |
+
"step": 10500
|
| 178 |
+
},
|
| 179 |
+
{
|
| 180 |
+
"epoch": 0.14945956779010441,
|
| 181 |
+
"grad_norm": 1.3417694568634033,
|
| 182 |
+
"learning_rate": 0.00042538520068480126,
|
| 183 |
+
"loss": 1.7324,
|
| 184 |
+
"mean_token_accuracy": 0.5957142559885978,
|
| 185 |
+
"step": 11000
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.15625318450783643,
|
| 189 |
+
"grad_norm": 1.3369439840316772,
|
| 190 |
+
"learning_rate": 0.00042200195657490693,
|
| 191 |
+
"loss": 1.7428,
|
| 192 |
+
"mean_token_accuracy": 0.5911534672677516,
|
| 193 |
+
"step": 11500
|
| 194 |
+
},
|
| 195 |
+
{
|
| 196 |
+
"epoch": 0.16304680122556844,
|
| 197 |
+
"grad_norm": 1.2490959167480469,
|
| 198 |
+
"learning_rate": 0.00041860512513927005,
|
| 199 |
+
"loss": 1.7318,
|
| 200 |
+
"mean_token_accuracy": 0.5937872271835803,
|
| 201 |
+
"step": 12000
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"epoch": 0.1698404179433005,
|
| 205 |
+
"grad_norm": 1.2229673862457275,
|
| 206 |
+
"learning_rate": 0.0004152218810293758,
|
| 207 |
+
"loss": 1.7429,
|
| 208 |
+
"mean_token_accuracy": 0.5896593699157238,
|
| 209 |
+
"step": 12500
|
| 210 |
+
},
|
| 211 |
+
{
|
| 212 |
+
"epoch": 0.1766340346610325,
|
| 213 |
+
"grad_norm": 1.3295321464538574,
|
| 214 |
+
"learning_rate": 0.00041182504959373894,
|
| 215 |
+
"loss": 1.7633,
|
| 216 |
+
"mean_token_accuracy": 0.5877331503927707,
|
| 217 |
+
"step": 13000
|
| 218 |
+
},
|
| 219 |
+
{
|
| 220 |
+
"epoch": 0.18342765137876452,
|
| 221 |
+
"grad_norm": 0.9519675970077515,
|
| 222 |
+
"learning_rate": 0.00040844180548384466,
|
| 223 |
+
"loss": 1.7573,
|
| 224 |
+
"mean_token_accuracy": 0.5893017112016677,
|
| 225 |
+
"step": 13500
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.19022126809649653,
|
| 229 |
+
"grad_norm": 1.448428750038147,
|
| 230 |
+
"learning_rate": 0.00040504497404820783,
|
| 231 |
+
"loss": 1.7314,
|
| 232 |
+
"mean_token_accuracy": 0.5936685436069965,
|
| 233 |
+
"step": 14000
|
| 234 |
+
},
|
| 235 |
+
{
|
| 236 |
+
"epoch": 0.19701488481422855,
|
| 237 |
+
"grad_norm": 1.3305081129074097,
|
| 238 |
+
"learning_rate": 0.00040166172993831356,
|
| 239 |
+
"loss": 1.7293,
|
| 240 |
+
"mean_token_accuracy": 0.5929152786135673,
|
| 241 |
+
"step": 14500
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.20380850153196056,
|
| 245 |
+
"grad_norm": 1.2279210090637207,
|
| 246 |
+
"learning_rate": 0.00039826489850267673,
|
| 247 |
+
"loss": 1.728,
|
| 248 |
+
"mean_token_accuracy": 0.5941076972186565,
|
| 249 |
+
"step": 15000
|
| 250 |
+
},
|
| 251 |
+
{
|
| 252 |
+
"epoch": 0.21060211824969258,
|
| 253 |
+
"grad_norm": 0.9746572375297546,
|
| 254 |
+
"learning_rate": 0.0003948816543927824,
|
| 255 |
+
"loss": 1.7452,
|
| 256 |
+
"mean_token_accuracy": 0.5927870056033134,
|
| 257 |
+
"step": 15500
|
| 258 |
+
},
|
| 259 |
+
{
|
| 260 |
+
"epoch": 0.21739573496742462,
|
| 261 |
+
"grad_norm": 1.3720732927322388,
|
| 262 |
+
"learning_rate": 0.0003914848229571456,
|
| 263 |
+
"loss": 1.7259,
|
| 264 |
+
"mean_token_accuracy": 0.5940872375369072,
|
| 265 |
+
"step": 16000
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.22418935168515663,
|
| 269 |
+
"grad_norm": 1.2336689233779907,
|
| 270 |
+
"learning_rate": 0.0003881015788472513,
|
| 271 |
+
"loss": 1.7365,
|
| 272 |
+
"mean_token_accuracy": 0.5929380969107151,
|
| 273 |
+
"step": 16500
|
| 274 |
+
},
|
| 275 |
+
{
|
| 276 |
+
"epoch": 0.23098296840288865,
|
| 277 |
+
"grad_norm": 1.0067358016967773,
|
| 278 |
+
"learning_rate": 0.0003847047474116144,
|
| 279 |
+
"loss": 1.7373,
|
| 280 |
+
"mean_token_accuracy": 0.5914620462656021,
|
| 281 |
+
"step": 17000
|
| 282 |
+
},
|
| 283 |
+
{
|
| 284 |
+
"epoch": 0.23777658512062066,
|
| 285 |
+
"grad_norm": 1.2364681959152222,
|
| 286 |
+
"learning_rate": 0.0003813215033017202,
|
| 287 |
+
"loss": 1.7298,
|
| 288 |
+
"mean_token_accuracy": 0.5921443883776665,
|
| 289 |
+
"step": 17500
|
| 290 |
+
},
|
| 291 |
+
{
|
| 292 |
+
"epoch": 0.24457020183835268,
|
| 293 |
+
"grad_norm": 1.186237096786499,
|
| 294 |
+
"learning_rate": 0.0003779246718660833,
|
| 295 |
+
"loss": 1.7336,
|
| 296 |
+
"mean_token_accuracy": 0.5918942322134971,
|
| 297 |
+
"step": 18000
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.2513638185560847,
|
| 301 |
+
"grad_norm": 1.2172770500183105,
|
| 302 |
+
"learning_rate": 0.000374541427756189,
|
| 303 |
+
"loss": 1.7321,
|
| 304 |
+
"mean_token_accuracy": 0.592568263143301,
|
| 305 |
+
"step": 18500
|
| 306 |
+
},
|
| 307 |
+
{
|
| 308 |
+
"epoch": 0.25815743527381674,
|
| 309 |
+
"grad_norm": 1.074506402015686,
|
| 310 |
+
"learning_rate": 0.0003711445963205522,
|
| 311 |
+
"loss": 1.726,
|
| 312 |
+
"mean_token_accuracy": 0.5930768192112446,
|
| 313 |
+
"step": 19000
|
| 314 |
+
},
|
| 315 |
+
{
|
| 316 |
+
"epoch": 0.2649510519915487,
|
| 317 |
+
"grad_norm": 1.3433698415756226,
|
| 318 |
+
"learning_rate": 0.0003677681458735292,
|
| 319 |
+
"loss": 1.7252,
|
| 320 |
+
"mean_token_accuracy": 0.5934848180115223,
|
| 321 |
+
"step": 19500
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.27174466870928077,
|
| 325 |
+
"grad_norm": 1.0935838222503662,
|
| 326 |
+
"learning_rate": 0.00036437131443789233,
|
| 327 |
+
"loss": 1.7276,
|
| 328 |
+
"mean_token_accuracy": 0.5916450168192386,
|
| 329 |
+
"step": 20000
|
| 330 |
+
},
|
| 331 |
+
{
|
| 332 |
+
"epoch": 0.27853828542701276,
|
| 333 |
+
"grad_norm": 1.5121397972106934,
|
| 334 |
+
"learning_rate": 0.0003609744830022555,
|
| 335 |
+
"loss": 1.738,
|
| 336 |
+
"mean_token_accuracy": 0.5906197059154511,
|
| 337 |
+
"step": 20500
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"epoch": 0.2853319021447448,
|
| 341 |
+
"grad_norm": 1.2155652046203613,
|
| 342 |
+
"learning_rate": 0.0003575776515666186,
|
| 343 |
+
"loss": 1.7221,
|
| 344 |
+
"mean_token_accuracy": 0.5948817262351513,
|
| 345 |
+
"step": 21000
|
| 346 |
+
},
|
| 347 |
+
{
|
| 348 |
+
"epoch": 0.29212551886247684,
|
| 349 |
+
"grad_norm": 1.0498167276382446,
|
| 350 |
+
"learning_rate": 0.0003541944074567244,
|
| 351 |
+
"loss": 1.7346,
|
| 352 |
+
"mean_token_accuracy": 0.5901701744496822,
|
| 353 |
+
"step": 21500
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.29891913558020883,
|
| 357 |
+
"grad_norm": 0.9237672686576843,
|
| 358 |
+
"learning_rate": 0.0003507975760210875,
|
| 359 |
+
"loss": 1.7265,
|
| 360 |
+
"mean_token_accuracy": 0.5947049247324466,
|
| 361 |
+
"step": 22000
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.30571275229794087,
|
| 365 |
+
"grad_norm": 1.235968828201294,
|
| 366 |
+
"learning_rate": 0.00034741433191119324,
|
| 367 |
+
"loss": 1.7359,
|
| 368 |
+
"mean_token_accuracy": 0.5912741076052189,
|
| 369 |
+
"step": 22500
|
| 370 |
+
},
|
| 371 |
+
{
|
| 372 |
+
"epoch": 0.31250636901567286,
|
| 373 |
+
"grad_norm": 1.3003114461898804,
|
| 374 |
+
"learning_rate": 0.0003440175004755564,
|
| 375 |
+
"loss": 1.7165,
|
| 376 |
+
"mean_token_accuracy": 0.596044946283102,
|
| 377 |
+
"step": 23000
|
| 378 |
+
},
|
| 379 |
+
{
|
| 380 |
+
"epoch": 0.3192999857334049,
|
| 381 |
+
"grad_norm": 0.9103087186813354,
|
| 382 |
+
"learning_rate": 0.00034063425636566213,
|
| 383 |
+
"loss": 1.7145,
|
| 384 |
+
"mean_token_accuracy": 0.5962611735165119,
|
| 385 |
+
"step": 23500
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.3260936024511369,
|
| 389 |
+
"grad_norm": 1.3072879314422607,
|
| 390 |
+
"learning_rate": 0.0003372374249300253,
|
| 391 |
+
"loss": 1.7182,
|
| 392 |
+
"mean_token_accuracy": 0.5958844081163407,
|
| 393 |
+
"step": 24000
|
| 394 |
+
},
|
| 395 |
+
{
|
| 396 |
+
"epoch": 0.33288721916886893,
|
| 397 |
+
"grad_norm": 1.122262716293335,
|
| 398 |
+
"learning_rate": 0.00033385418082013097,
|
| 399 |
+
"loss": 1.7277,
|
| 400 |
+
"mean_token_accuracy": 0.5924356364905834,
|
| 401 |
+
"step": 24500
|
| 402 |
+
},
|
| 403 |
+
{
|
| 404 |
+
"epoch": 0.339680835886601,
|
| 405 |
+
"grad_norm": 1.0892757177352905,
|
| 406 |
+
"learning_rate": 0.0003304573493844942,
|
| 407 |
+
"loss": 1.7168,
|
| 408 |
+
"mean_token_accuracy": 0.5951408152878285,
|
| 409 |
+
"step": 25000
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.34647445260433296,
|
| 413 |
+
"grad_norm": 1.3095890283584595,
|
| 414 |
+
"learning_rate": 0.00032707410527459986,
|
| 415 |
+
"loss": 1.7103,
|
| 416 |
+
"mean_token_accuracy": 0.5958214187920093,
|
| 417 |
+
"step": 25500
|
| 418 |
+
},
|
| 419 |
+
{
|
| 420 |
+
"epoch": 0.353268069322065,
|
| 421 |
+
"grad_norm": 0.9692500233650208,
|
| 422 |
+
"learning_rate": 0.000323677273838963,
|
| 423 |
+
"loss": 1.7127,
|
| 424 |
+
"mean_token_accuracy": 0.5952938674986362,
|
| 425 |
+
"step": 26000
|
| 426 |
+
},
|
| 427 |
+
{
|
| 428 |
+
"epoch": 0.360061686039797,
|
| 429 |
+
"grad_norm": 1.0970525741577148,
|
| 430 |
+
"learning_rate": 0.00032029402972906876,
|
| 431 |
+
"loss": 1.7251,
|
| 432 |
+
"mean_token_accuracy": 0.5936492686867714,
|
| 433 |
+
"step": 26500
|
| 434 |
+
},
|
| 435 |
+
{
|
| 436 |
+
"epoch": 0.36685530275752903,
|
| 437 |
+
"grad_norm": 1.2200478315353394,
|
| 438 |
+
"learning_rate": 0.00031689719829343187,
|
| 439 |
+
"loss": 1.7082,
|
| 440 |
+
"mean_token_accuracy": 0.5961000478565693,
|
| 441 |
+
"step": 27000
|
| 442 |
+
},
|
| 443 |
+
{
|
| 444 |
+
"epoch": 0.373648919475261,
|
| 445 |
+
"grad_norm": 1.0182169675827026,
|
| 446 |
+
"learning_rate": 0.0003135139541835376,
|
| 447 |
+
"loss": 1.7038,
|
| 448 |
+
"mean_token_accuracy": 0.5973086108565331,
|
| 449 |
+
"step": 27500
|
| 450 |
+
},
|
| 451 |
+
{
|
| 452 |
+
"epoch": 0.38044253619299306,
|
| 453 |
+
"grad_norm": 0.9256690740585327,
|
| 454 |
+
"learning_rate": 0.00031011712274790076,
|
| 455 |
+
"loss": 1.7208,
|
| 456 |
+
"mean_token_accuracy": 0.593766670614481,
|
| 457 |
+
"step": 28000
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.3872361529107251,
|
| 461 |
+
"grad_norm": 1.1710305213928223,
|
| 462 |
+
"learning_rate": 0.0003067270849751352,
|
| 463 |
+
"loss": 1.73,
|
| 464 |
+
"mean_token_accuracy": 0.594021442502737,
|
| 465 |
+
"step": 28500
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.3940297696284571,
|
| 469 |
+
"grad_norm": 1.1612814664840698,
|
| 470 |
+
"learning_rate": 0.00030333704720236966,
|
| 471 |
+
"loss": 1.7146,
|
| 472 |
+
"mean_token_accuracy": 0.5956000906527043,
|
| 473 |
+
"step": 29000
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"epoch": 0.40082338634618914,
|
| 477 |
+
"grad_norm": 1.2125431299209595,
|
| 478 |
+
"learning_rate": 0.00029994021576673283,
|
| 479 |
+
"loss": 1.7076,
|
| 480 |
+
"mean_token_accuracy": 0.5957852031290531,
|
| 481 |
+
"step": 29500
|
| 482 |
+
},
|
| 483 |
+
{
|
| 484 |
+
"epoch": 0.4076170030639211,
|
| 485 |
+
"grad_norm": 1.2176785469055176,
|
| 486 |
+
"learning_rate": 0.0002965569716568385,
|
| 487 |
+
"loss": 1.7312,
|
| 488 |
+
"mean_token_accuracy": 0.5945550291240216,
|
| 489 |
+
"step": 30000
|
| 490 |
+
},
|
| 491 |
+
{
|
| 492 |
+
"epoch": 0.41441061978165317,
|
| 493 |
+
"grad_norm": 1.5686943531036377,
|
| 494 |
+
"learning_rate": 0.00029316014022120167,
|
| 495 |
+
"loss": 1.7171,
|
| 496 |
+
"mean_token_accuracy": 0.5942135262489319,
|
| 497 |
+
"step": 30500
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.42120423649938515,
|
| 501 |
+
"grad_norm": 1.0116132497787476,
|
| 502 |
+
"learning_rate": 0.0002897768961113074,
|
| 503 |
+
"loss": 1.6985,
|
| 504 |
+
"mean_token_accuracy": 0.5982498173713684,
|
| 505 |
+
"step": 31000
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.4279978532171172,
|
| 509 |
+
"grad_norm": 0.9668023586273193,
|
| 510 |
+
"learning_rate": 0.0002863800646756705,
|
| 511 |
+
"loss": 1.7078,
|
| 512 |
+
"mean_token_accuracy": 0.5950165801644325,
|
| 513 |
+
"step": 31500
|
| 514 |
+
},
|
| 515 |
+
{
|
| 516 |
+
"epoch": 0.43479146993484924,
|
| 517 |
+
"grad_norm": 1.2987300157546997,
|
| 518 |
+
"learning_rate": 0.00028299682056577623,
|
| 519 |
+
"loss": 1.7288,
|
| 520 |
+
"mean_token_accuracy": 0.5917907827198505,
|
| 521 |
+
"step": 32000
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.4415850866525812,
|
| 525 |
+
"grad_norm": 1.1014894247055054,
|
| 526 |
+
"learning_rate": 0.0002795999891301394,
|
| 527 |
+
"loss": 1.7207,
|
| 528 |
+
"mean_token_accuracy": 0.5922197285294533,
|
| 529 |
+
"step": 32500
|
| 530 |
+
},
|
| 531 |
+
{
|
| 532 |
+
"epoch": 0.44837870337031327,
|
| 533 |
+
"grad_norm": 1.0592535734176636,
|
| 534 |
+
"learning_rate": 0.0002762167450202451,
|
| 535 |
+
"loss": 1.7043,
|
| 536 |
+
"mean_token_accuracy": 0.5975054582953453,
|
| 537 |
+
"step": 33000
|
| 538 |
+
},
|
| 539 |
+
{
|
| 540 |
+
"epoch": 0.45517232008804526,
|
| 541 |
+
"grad_norm": 1.084416151046753,
|
| 542 |
+
"learning_rate": 0.0002728199135846083,
|
| 543 |
+
"loss": 1.7153,
|
| 544 |
+
"mean_token_accuracy": 0.59563816472888,
|
| 545 |
+
"step": 33500
|
| 546 |
+
},
|
| 547 |
+
{
|
| 548 |
+
"epoch": 0.4619659368057773,
|
| 549 |
+
"grad_norm": 1.0490503311157227,
|
| 550 |
+
"learning_rate": 0.000269436669474714,
|
| 551 |
+
"loss": 1.7049,
|
| 552 |
+
"mean_token_accuracy": 0.5960744704604148,
|
| 553 |
+
"step": 34000
|
| 554 |
+
},
|
| 555 |
+
{
|
| 556 |
+
"epoch": 0.4687595535235093,
|
| 557 |
+
"grad_norm": 1.3198046684265137,
|
| 558 |
+
"learning_rate": 0.0002660398380390772,
|
| 559 |
+
"loss": 1.6951,
|
| 560 |
+
"mean_token_accuracy": 0.6000102725625038,
|
| 561 |
+
"step": 34500
|
| 562 |
+
},
|
| 563 |
+
{
|
| 564 |
+
"epoch": 0.47555317024124133,
|
| 565 |
+
"grad_norm": 1.1859066486358643,
|
| 566 |
+
"learning_rate": 0.00026265659392918285,
|
| 567 |
+
"loss": 1.7025,
|
| 568 |
+
"mean_token_accuracy": 0.5961538652777671,
|
| 569 |
+
"step": 35000
|
| 570 |
+
},
|
| 571 |
+
{
|
| 572 |
+
"epoch": 0.4823467869589734,
|
| 573 |
+
"grad_norm": 1.0394221544265747,
|
| 574 |
+
"learning_rate": 0.000259259762493546,
|
| 575 |
+
"loss": 1.695,
|
| 576 |
+
"mean_token_accuracy": 0.599614026337862,
|
| 577 |
+
"step": 35500
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.48914040367670536,
|
| 581 |
+
"grad_norm": 1.1829257011413574,
|
| 582 |
+
"learning_rate": 0.00025587651838365175,
|
| 583 |
+
"loss": 1.7086,
|
| 584 |
+
"mean_token_accuracy": 0.5964278879761696,
|
| 585 |
+
"step": 36000
|
| 586 |
+
},
|
| 587 |
+
{
|
| 588 |
+
"epoch": 0.4959340203944374,
|
| 589 |
+
"grad_norm": 1.0766243934631348,
|
| 590 |
+
"learning_rate": 0.00025247968694801486,
|
| 591 |
+
"loss": 1.7,
|
| 592 |
+
"mean_token_accuracy": 0.5977904357016086,
|
| 593 |
+
"step": 36500
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.5027276371121694,
|
| 597 |
+
"grad_norm": 1.1784560680389404,
|
| 598 |
+
"learning_rate": 0.0002490964428381206,
|
| 599 |
+
"loss": 1.715,
|
| 600 |
+
"mean_token_accuracy": 0.5959987764656544,
|
| 601 |
+
"step": 37000
|
| 602 |
+
},
|
| 603 |
+
{
|
| 604 |
+
"epoch": 0.5095212538299014,
|
| 605 |
+
"grad_norm": 1.1208888292312622,
|
| 606 |
+
"learning_rate": 0.00024569961140248376,
|
| 607 |
+
"loss": 1.7046,
|
| 608 |
+
"mean_token_accuracy": 0.5964538981616497,
|
| 609 |
+
"step": 37500
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"epoch": 0.5163148705476335,
|
| 613 |
+
"grad_norm": 1.0677920579910278,
|
| 614 |
+
"learning_rate": 0.00024231636729258948,
|
| 615 |
+
"loss": 1.7001,
|
| 616 |
+
"mean_token_accuracy": 0.5974933596849441,
|
| 617 |
+
"step": 38000
|
| 618 |
+
},
|
| 619 |
+
{
|
| 620 |
+
"epoch": 0.5231084872653654,
|
| 621 |
+
"grad_norm": 0.9215283393859863,
|
| 622 |
+
"learning_rate": 0.00023891953585695265,
|
| 623 |
+
"loss": 1.7105,
|
| 624 |
+
"mean_token_accuracy": 0.596855210095644,
|
| 625 |
+
"step": 38500
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.5299021039830974,
|
| 629 |
+
"grad_norm": 1.23247492313385,
|
| 630 |
+
"learning_rate": 0.00023553629174705835,
|
| 631 |
+
"loss": 1.711,
|
| 632 |
+
"mean_token_accuracy": 0.5971770713925362,
|
| 633 |
+
"step": 39000
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.5366957207008295,
|
| 637 |
+
"grad_norm": 1.07834792137146,
|
| 638 |
+
"learning_rate": 0.00023213946031142152,
|
| 639 |
+
"loss": 1.6993,
|
| 640 |
+
"mean_token_accuracy": 0.5984128168821334,
|
| 641 |
+
"step": 39500
|
| 642 |
+
},
|
| 643 |
+
{
|
| 644 |
+
"epoch": 0.5434893374185615,
|
| 645 |
+
"grad_norm": 1.1771501302719116,
|
| 646 |
+
"learning_rate": 0.00022875621620152724,
|
| 647 |
+
"loss": 1.6937,
|
| 648 |
+
"mean_token_accuracy": 0.6004006116986275,
|
| 649 |
+
"step": 40000
|
| 650 |
+
},
|
| 651 |
+
{
|
| 652 |
+
"epoch": 0.5502829541362936,
|
| 653 |
+
"grad_norm": 1.2967567443847656,
|
| 654 |
+
"learning_rate": 0.00022535938476589038,
|
| 655 |
+
"loss": 1.7038,
|
| 656 |
+
"mean_token_accuracy": 0.5961601254045963,
|
| 657 |
+
"step": 40500
|
| 658 |
+
},
|
| 659 |
+
{
|
| 660 |
+
"epoch": 0.5570765708540255,
|
| 661 |
+
"grad_norm": 1.1175864934921265,
|
| 662 |
+
"learning_rate": 0.00022197614065599608,
|
| 663 |
+
"loss": 1.716,
|
| 664 |
+
"mean_token_accuracy": 0.5941783719956875,
|
| 665 |
+
"step": 41000
|
| 666 |
+
},
|
| 667 |
+
{
|
| 668 |
+
"epoch": 0.5638701875717576,
|
| 669 |
+
"grad_norm": 1.380516529083252,
|
| 670 |
+
"learning_rate": 0.00021857930922035927,
|
| 671 |
+
"loss": 1.6936,
|
| 672 |
+
"mean_token_accuracy": 0.5974556840360165,
|
| 673 |
+
"step": 41500
|
| 674 |
+
},
|
| 675 |
+
{
|
| 676 |
+
"epoch": 0.5706638042894896,
|
| 677 |
+
"grad_norm": 1.1645792722702026,
|
| 678 |
+
"learning_rate": 0.00021519606511046497,
|
| 679 |
+
"loss": 1.6964,
|
| 680 |
+
"mean_token_accuracy": 0.5985994503200054,
|
| 681 |
+
"step": 42000
|
| 682 |
+
},
|
| 683 |
+
{
|
| 684 |
+
"epoch": 0.5774574210072216,
|
| 685 |
+
"grad_norm": 1.3061178922653198,
|
| 686 |
+
"learning_rate": 0.0002117992336748281,
|
| 687 |
+
"loss": 1.7065,
|
| 688 |
+
"mean_token_accuracy": 0.5961524350643158,
|
| 689 |
+
"step": 42500
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.5842510377249537,
|
| 693 |
+
"grad_norm": 1.1523349285125732,
|
| 694 |
+
"learning_rate": 0.0002084159895649338,
|
| 695 |
+
"loss": 1.7001,
|
| 696 |
+
"mean_token_accuracy": 0.5993789280354976,
|
| 697 |
+
"step": 43000
|
| 698 |
+
},
|
| 699 |
+
{
|
| 700 |
+
"epoch": 0.5910446544426856,
|
| 701 |
+
"grad_norm": 1.1238070726394653,
|
| 702 |
+
"learning_rate": 0.000205019158129297,
|
| 703 |
+
"loss": 1.724,
|
| 704 |
+
"mean_token_accuracy": 0.5929664156734943,
|
| 705 |
+
"step": 43500
|
| 706 |
+
},
|
| 707 |
+
{
|
| 708 |
+
"epoch": 0.5978382711604177,
|
| 709 |
+
"grad_norm": 1.1592812538146973,
|
| 710 |
+
"learning_rate": 0.0002016359140194027,
|
| 711 |
+
"loss": 1.7065,
|
| 712 |
+
"mean_token_accuracy": 0.5956415718495846,
|
| 713 |
+
"step": 44000
|
| 714 |
+
},
|
| 715 |
+
{
|
| 716 |
+
"epoch": 0.6046318878781497,
|
| 717 |
+
"grad_norm": 1.080269694328308,
|
| 718 |
+
"learning_rate": 0.00019823908258376585,
|
| 719 |
+
"loss": 1.6983,
|
| 720 |
+
"mean_token_accuracy": 0.5975038340389729,
|
| 721 |
+
"step": 44500
|
| 722 |
+
},
|
| 723 |
+
{
|
| 724 |
+
"epoch": 0.6114255045958817,
|
| 725 |
+
"grad_norm": 1.281002163887024,
|
| 726 |
+
"learning_rate": 0.0001948558384738716,
|
| 727 |
+
"loss": 1.6965,
|
| 728 |
+
"mean_token_accuracy": 0.5972202825844288,
|
| 729 |
+
"step": 45000
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.6182191213136137,
|
| 733 |
+
"grad_norm": 1.004654049873352,
|
| 734 |
+
"learning_rate": 0.00019145900703823474,
|
| 735 |
+
"loss": 1.7176,
|
| 736 |
+
"mean_token_accuracy": 0.5956167621910572,
|
| 737 |
+
"step": 45500
|
| 738 |
+
},
|
| 739 |
+
{
|
| 740 |
+
"epoch": 0.6250127380313457,
|
| 741 |
+
"grad_norm": 1.0295976400375366,
|
| 742 |
+
"learning_rate": 0.00018807576292834043,
|
| 743 |
+
"loss": 1.6929,
|
| 744 |
+
"mean_token_accuracy": 0.5980769891738892,
|
| 745 |
+
"step": 46000
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"epoch": 0.6318063547490778,
|
| 749 |
+
"grad_norm": 1.0710487365722656,
|
| 750 |
+
"learning_rate": 0.00018467893149270363,
|
| 751 |
+
"loss": 1.7116,
|
| 752 |
+
"mean_token_accuracy": 0.5948944690227509,
|
| 753 |
+
"step": 46500
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.6385999714668098,
|
| 757 |
+
"grad_norm": 1.1924978494644165,
|
| 758 |
+
"learning_rate": 0.00018129568738280933,
|
| 759 |
+
"loss": 1.6896,
|
| 760 |
+
"mean_token_accuracy": 0.6006084454953671,
|
| 761 |
+
"step": 47000
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 0.6453935881845418,
|
| 765 |
+
"grad_norm": 1.1349835395812988,
|
| 766 |
+
"learning_rate": 0.00017789885594717247,
|
| 767 |
+
"loss": 1.706,
|
| 768 |
+
"mean_token_accuracy": 0.5981652895510197,
|
| 769 |
+
"step": 47500
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.6521872049022738,
|
| 773 |
+
"grad_norm": 1.4327572584152222,
|
| 774 |
+
"learning_rate": 0.0001745156118372782,
|
| 775 |
+
"loss": 1.7034,
|
| 776 |
+
"mean_token_accuracy": 0.5959363037645817,
|
| 777 |
+
"step": 48000
|
| 778 |
+
},
|
| 779 |
+
{
|
| 780 |
+
"epoch": 0.6589808216200058,
|
| 781 |
+
"grad_norm": 1.499800682067871,
|
| 782 |
+
"learning_rate": 0.00017111878040164136,
|
| 783 |
+
"loss": 1.696,
|
| 784 |
+
"mean_token_accuracy": 0.598519076347351,
|
| 785 |
+
"step": 48500
|
| 786 |
+
},
|
| 787 |
+
{
|
| 788 |
+
"epoch": 0.6657744383377379,
|
| 789 |
+
"grad_norm": 1.173693299293518,
|
| 790 |
+
"learning_rate": 0.00016773553629174706,
|
| 791 |
+
"loss": 1.6872,
|
| 792 |
+
"mean_token_accuracy": 0.6003931475877762,
|
| 793 |
+
"step": 49000
|
| 794 |
+
},
|
| 795 |
+
{
|
| 796 |
+
"epoch": 0.6725680550554699,
|
| 797 |
+
"grad_norm": 1.0841670036315918,
|
| 798 |
+
"learning_rate": 0.0001643387048561102,
|
| 799 |
+
"loss": 1.7033,
|
| 800 |
+
"mean_token_accuracy": 0.5981890520751476,
|
| 801 |
+
"step": 49500
|
| 802 |
+
},
|
| 803 |
+
{
|
| 804 |
+
"epoch": 0.679361671773202,
|
| 805 |
+
"grad_norm": 1.2305493354797363,
|
| 806 |
+
"learning_rate": 0.00016095546074621593,
|
| 807 |
+
"loss": 1.7026,
|
| 808 |
+
"mean_token_accuracy": 0.5967764587402343,
|
| 809 |
+
"step": 50000
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.6861552884909339,
|
| 813 |
+
"grad_norm": 1.1358468532562256,
|
| 814 |
+
"learning_rate": 0.0001575586293105791,
|
| 815 |
+
"loss": 1.7024,
|
| 816 |
+
"mean_token_accuracy": 0.5969898876547813,
|
| 817 |
+
"step": 50500
|
| 818 |
+
},
|
| 819 |
+
{
|
| 820 |
+
"epoch": 0.6929489052086659,
|
| 821 |
+
"grad_norm": 1.3159434795379639,
|
| 822 |
+
"learning_rate": 0.0001541753852006848,
|
| 823 |
+
"loss": 1.6995,
|
| 824 |
+
"mean_token_accuracy": 0.5962181152105331,
|
| 825 |
+
"step": 51000
|
| 826 |
+
},
|
| 827 |
+
{
|
| 828 |
+
"epoch": 0.699742521926398,
|
| 829 |
+
"grad_norm": 1.27256441116333,
|
| 830 |
+
"learning_rate": 0.00015077855376504796,
|
| 831 |
+
"loss": 1.7029,
|
| 832 |
+
"mean_token_accuracy": 0.5967380773425103,
|
| 833 |
+
"step": 51500
|
| 834 |
+
},
|
| 835 |
+
{
|
| 836 |
+
"epoch": 0.70653613864413,
|
| 837 |
+
"grad_norm": 1.5178892612457275,
|
| 838 |
+
"learning_rate": 0.00014739530965515369,
|
| 839 |
+
"loss": 1.6899,
|
| 840 |
+
"mean_token_accuracy": 0.5981109066009521,
|
| 841 |
+
"step": 52000
|
| 842 |
+
},
|
| 843 |
+
{
|
| 844 |
+
"epoch": 0.7133297553618619,
|
| 845 |
+
"grad_norm": 0.9817212820053101,
|
| 846 |
+
"learning_rate": 0.00014399847821951683,
|
| 847 |
+
"loss": 1.6943,
|
| 848 |
+
"mean_token_accuracy": 0.5980751592516899,
|
| 849 |
+
"step": 52500
|
| 850 |
+
},
|
| 851 |
+
{
|
| 852 |
+
"epoch": 0.720123372079594,
|
| 853 |
+
"grad_norm": 1.0862836837768555,
|
| 854 |
+
"learning_rate": 0.00014061523410962255,
|
| 855 |
+
"loss": 1.7085,
|
| 856 |
+
"mean_token_accuracy": 0.5958100288212299,
|
| 857 |
+
"step": 53000
|
| 858 |
+
},
|
| 859 |
+
{
|
| 860 |
+
"epoch": 0.726916988797326,
|
| 861 |
+
"grad_norm": 1.3386940956115723,
|
| 862 |
+
"learning_rate": 0.00013721840267398572,
|
| 863 |
+
"loss": 1.6903,
|
| 864 |
+
"mean_token_accuracy": 0.5972609171569347,
|
| 865 |
+
"step": 53500
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.7337106055150581,
|
| 869 |
+
"grad_norm": 1.6551252603530884,
|
| 870 |
+
"learning_rate": 0.00013383515856409142,
|
| 871 |
+
"loss": 1.6976,
|
| 872 |
+
"mean_token_accuracy": 0.5984482218325138,
|
| 873 |
+
"step": 54000
|
| 874 |
+
},
|
| 875 |
+
{
|
| 876 |
+
"epoch": 0.7405042222327901,
|
| 877 |
+
"grad_norm": 1.1888887882232666,
|
| 878 |
+
"learning_rate": 0.0001304383271284546,
|
| 879 |
+
"loss": 1.7019,
|
| 880 |
+
"mean_token_accuracy": 0.5954872596561909,
|
| 881 |
+
"step": 54500
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 0.747297838950522,
|
| 885 |
+
"grad_norm": 1.0421268939971924,
|
| 886 |
+
"learning_rate": 0.00012705508301856028,
|
| 887 |
+
"loss": 1.6823,
|
| 888 |
+
"mean_token_accuracy": 0.5998311370313167,
|
| 889 |
+
"step": 55000
|
| 890 |
+
},
|
| 891 |
+
{
|
| 892 |
+
"epoch": 0.7540914556682541,
|
| 893 |
+
"grad_norm": 1.0341774225234985,
|
| 894 |
+
"learning_rate": 0.00012365825158292345,
|
| 895 |
+
"loss": 1.7042,
|
| 896 |
+
"mean_token_accuracy": 0.5982704365849495,
|
| 897 |
+
"step": 55500
|
| 898 |
+
},
|
| 899 |
+
{
|
| 900 |
+
"epoch": 0.7608850723859861,
|
| 901 |
+
"grad_norm": 1.222990870475769,
|
| 902 |
+
"learning_rate": 0.00012027500747302916,
|
| 903 |
+
"loss": 1.6795,
|
| 904 |
+
"mean_token_accuracy": 0.6036300667524338,
|
| 905 |
+
"step": 56000
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.7676786891037182,
|
| 909 |
+
"grad_norm": 1.0490987300872803,
|
| 910 |
+
"learning_rate": 0.00011687817603739233,
|
| 911 |
+
"loss": 1.6875,
|
| 912 |
+
"mean_token_accuracy": 0.597304708212614,
|
| 913 |
+
"step": 56500
|
| 914 |
+
},
|
| 915 |
+
{
|
| 916 |
+
"epoch": 0.7744723058214502,
|
| 917 |
+
"grad_norm": 1.1952784061431885,
|
| 918 |
+
"learning_rate": 0.00011349493192749803,
|
| 919 |
+
"loss": 1.6796,
|
| 920 |
+
"mean_token_accuracy": 0.6028886337280274,
|
| 921 |
+
"step": 57000
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.7812659225391821,
|
| 925 |
+
"grad_norm": 1.000869870185852,
|
| 926 |
+
"learning_rate": 0.0001100981004918612,
|
| 927 |
+
"loss": 1.7052,
|
| 928 |
+
"mean_token_accuracy": 0.5968053241670132,
|
| 929 |
+
"step": 57500
|
| 930 |
+
},
|
| 931 |
+
{
|
| 932 |
+
"epoch": 0.7880595392569142,
|
| 933 |
+
"grad_norm": 1.133309006690979,
|
| 934 |
+
"learning_rate": 0.0001067148563819669,
|
| 935 |
+
"loss": 1.6965,
|
| 936 |
+
"mean_token_accuracy": 0.5972890597879886,
|
| 937 |
+
"step": 58000
|
| 938 |
+
},
|
| 939 |
+
{
|
| 940 |
+
"epoch": 0.7948531559746462,
|
| 941 |
+
"grad_norm": 1.2667903900146484,
|
| 942 |
+
"learning_rate": 0.00010331802494633007,
|
| 943 |
+
"loss": 1.6786,
|
| 944 |
+
"mean_token_accuracy": 0.601085704267025,
|
| 945 |
+
"step": 58500
|
| 946 |
+
},
|
| 947 |
+
{
|
| 948 |
+
"epoch": 0.8016467726923783,
|
| 949 |
+
"grad_norm": 1.0539146661758423,
|
| 950 |
+
"learning_rate": 9.993478083643577e-05,
|
| 951 |
+
"loss": 1.7029,
|
| 952 |
+
"mean_token_accuracy": 0.5962156045734882,
|
| 953 |
+
"step": 59000
|
| 954 |
+
},
|
| 955 |
+
{
|
| 956 |
+
"epoch": 0.8084403894101102,
|
| 957 |
+
"grad_norm": 1.1023699045181274,
|
| 958 |
+
"learning_rate": 9.653794940079893e-05,
|
| 959 |
+
"loss": 1.6731,
|
| 960 |
+
"mean_token_accuracy": 0.6045135918557644,
|
| 961 |
+
"step": 59500
|
| 962 |
+
},
|
| 963 |
+
{
|
| 964 |
+
"epoch": 0.8152340061278422,
|
| 965 |
+
"grad_norm": 1.227386236190796,
|
| 966 |
+
"learning_rate": 9.315470529090464e-05,
|
| 967 |
+
"loss": 1.6855,
|
| 968 |
+
"mean_token_accuracy": 0.5978653418123722,
|
| 969 |
+
"step": 60000
|
| 970 |
+
},
|
| 971 |
+
{
|
| 972 |
+
"epoch": 0.8220276228455743,
|
| 973 |
+
"grad_norm": 1.2664766311645508,
|
| 974 |
+
"learning_rate": 8.976466751813909e-05,
|
| 975 |
+
"loss": 1.7055,
|
| 976 |
+
"mean_token_accuracy": 0.5958617155849933,
|
| 977 |
+
"step": 60500
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.8288212395633063,
|
| 981 |
+
"grad_norm": 1.263852834701538,
|
| 982 |
+
"learning_rate": 8.636783608250224e-05,
|
| 983 |
+
"loss": 1.6971,
|
| 984 |
+
"mean_token_accuracy": 0.5953974407315255,
|
| 985 |
+
"step": 61000
|
| 986 |
+
},
|
| 987 |
+
{
|
| 988 |
+
"epoch": 0.8356148562810384,
|
| 989 |
+
"grad_norm": 1.1483241319656372,
|
| 990 |
+
"learning_rate": 8.297100464686541e-05,
|
| 991 |
+
"loss": 1.6798,
|
| 992 |
+
"mean_token_accuracy": 0.5995199010372162,
|
| 993 |
+
"step": 61500
|
| 994 |
+
},
|
| 995 |
+
{
|
| 996 |
+
"epoch": 0.8424084729987703,
|
| 997 |
+
"grad_norm": 0.8838757872581482,
|
| 998 |
+
"learning_rate": 7.957417321122856e-05,
|
| 999 |
+
"loss": 1.6863,
|
| 1000 |
+
"mean_token_accuracy": 0.5994715365469456,
|
| 1001 |
+
"step": 62000
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.8492020897165024,
|
| 1005 |
+
"grad_norm": 1.0497251749038696,
|
| 1006 |
+
"learning_rate": 7.619092910133428e-05,
|
| 1007 |
+
"loss": 1.6692,
|
| 1008 |
+
"mean_token_accuracy": 0.6024055682122708,
|
| 1009 |
+
"step": 62500
|
| 1010 |
+
},
|
| 1011 |
+
{
|
| 1012 |
+
"epoch": 0.8559957064342344,
|
| 1013 |
+
"grad_norm": 0.9796432852745056,
|
| 1014 |
+
"learning_rate": 7.279409766569745e-05,
|
| 1015 |
+
"loss": 1.6799,
|
| 1016 |
+
"mean_token_accuracy": 0.5993573130667209,
|
| 1017 |
+
"step": 63000
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"epoch": 0.8627893231519664,
|
| 1021 |
+
"grad_norm": 1.2059897184371948,
|
| 1022 |
+
"learning_rate": 6.941085355580315e-05,
|
| 1023 |
+
"loss": 1.6763,
|
| 1024 |
+
"mean_token_accuracy": 0.6016550965607166,
|
| 1025 |
+
"step": 63500
|
| 1026 |
+
},
|
| 1027 |
+
{
|
| 1028 |
+
"epoch": 0.8695829398696985,
|
| 1029 |
+
"grad_norm": 1.184564232826233,
|
| 1030 |
+
"learning_rate": 6.601402212016632e-05,
|
| 1031 |
+
"loss": 1.683,
|
| 1032 |
+
"mean_token_accuracy": 0.5992586967945099,
|
| 1033 |
+
"step": 64000
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.8763765565874304,
|
| 1037 |
+
"grad_norm": 0.979015588760376,
|
| 1038 |
+
"learning_rate": 6.262398434740074e-05,
|
| 1039 |
+
"loss": 1.6916,
|
| 1040 |
+
"mean_token_accuracy": 0.6002075519263744,
|
| 1041 |
+
"step": 64500
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.8831701733051625,
|
| 1045 |
+
"grad_norm": 1.4841409921646118,
|
| 1046 |
+
"learning_rate": 5.924074023750645e-05,
|
| 1047 |
+
"loss": 1.6815,
|
| 1048 |
+
"mean_token_accuracy": 0.6010229328274727,
|
| 1049 |
+
"step": 65000
|
| 1050 |
+
},
|
| 1051 |
+
{
|
| 1052 |
+
"epoch": 0.8899637900228945,
|
| 1053 |
+
"grad_norm": 1.0104435682296753,
|
| 1054 |
+
"learning_rate": 5.5843908801869616e-05,
|
| 1055 |
+
"loss": 1.6942,
|
| 1056 |
+
"mean_token_accuracy": 0.5977232045531273,
|
| 1057 |
+
"step": 65500
|
| 1058 |
+
},
|
| 1059 |
+
{
|
| 1060 |
+
"epoch": 0.8967574067406265,
|
| 1061 |
+
"grad_norm": 1.364670991897583,
|
| 1062 |
+
"learning_rate": 5.244707736623278e-05,
|
| 1063 |
+
"loss": 1.6862,
|
| 1064 |
+
"mean_token_accuracy": 0.5978881956636906,
|
| 1065 |
+
"step": 66000
|
| 1066 |
+
},
|
| 1067 |
+
{
|
| 1068 |
+
"epoch": 0.9035510234583585,
|
| 1069 |
+
"grad_norm": 1.085087537765503,
|
| 1070 |
+
"learning_rate": 4.905024593059594e-05,
|
| 1071 |
+
"loss": 1.6812,
|
| 1072 |
+
"mean_token_accuracy": 0.5997030725479126,
|
| 1073 |
+
"step": 66500
|
| 1074 |
+
},
|
| 1075 |
+
{
|
| 1076 |
+
"epoch": 0.9103446401760905,
|
| 1077 |
+
"grad_norm": 0.9562463164329529,
|
| 1078 |
+
"learning_rate": 4.567379548357292e-05,
|
| 1079 |
+
"loss": 1.6848,
|
| 1080 |
+
"mean_token_accuracy": 0.5997031436264515,
|
| 1081 |
+
"step": 67000
|
| 1082 |
+
},
|
| 1083 |
+
{
|
| 1084 |
+
"epoch": 0.9171382568938226,
|
| 1085 |
+
"grad_norm": 1.2848726511001587,
|
| 1086 |
+
"learning_rate": 4.2276964047936085e-05,
|
| 1087 |
+
"loss": 1.6873,
|
| 1088 |
+
"mean_token_accuracy": 0.599891084432602,
|
| 1089 |
+
"step": 67500
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.9239318736115546,
|
| 1093 |
+
"grad_norm": 1.168323040008545,
|
| 1094 |
+
"learning_rate": 3.888013261229924e-05,
|
| 1095 |
+
"loss": 1.6807,
|
| 1096 |
+
"mean_token_accuracy": 0.599244915753603,
|
| 1097 |
+
"step": 68000
|
| 1098 |
+
},
|
| 1099 |
+
{
|
| 1100 |
+
"epoch": 0.9307254903292866,
|
| 1101 |
+
"grad_norm": 1.1764905452728271,
|
| 1102 |
+
"learning_rate": 3.548330117666241e-05,
|
| 1103 |
+
"loss": 1.6857,
|
| 1104 |
+
"mean_token_accuracy": 0.6015235537290573,
|
| 1105 |
+
"step": 68500
|
| 1106 |
+
},
|
| 1107 |
+
{
|
| 1108 |
+
"epoch": 0.9375191070470186,
|
| 1109 |
+
"grad_norm": 1.2431254386901855,
|
| 1110 |
+
"learning_rate": 3.210005706676812e-05,
|
| 1111 |
+
"loss": 1.6735,
|
| 1112 |
+
"mean_token_accuracy": 0.6007710628509522,
|
| 1113 |
+
"step": 69000
|
| 1114 |
+
},
|
| 1115 |
+
{
|
| 1116 |
+
"epoch": 0.9443127237647506,
|
| 1117 |
+
"grad_norm": 1.4310569763183594,
|
| 1118 |
+
"learning_rate": 2.870322563113128e-05,
|
| 1119 |
+
"loss": 1.6949,
|
| 1120 |
+
"mean_token_accuracy": 0.5974322560429574,
|
| 1121 |
+
"step": 69500
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 0.9511063404824827,
|
| 1125 |
+
"grad_norm": 1.0426503419876099,
|
| 1126 |
+
"learning_rate": 2.531998152123699e-05,
|
| 1127 |
+
"loss": 1.6717,
|
| 1128 |
+
"mean_token_accuracy": 0.6013675058782101,
|
| 1129 |
+
"step": 70000
|
| 1130 |
+
},
|
| 1131 |
+
{
|
| 1132 |
+
"epoch": 0.9578999572002147,
|
| 1133 |
+
"grad_norm": 1.1059454679489136,
|
| 1134 |
+
"learning_rate": 2.192315008560015e-05,
|
| 1135 |
+
"loss": 1.6838,
|
| 1136 |
+
"mean_token_accuracy": 0.5987102429866791,
|
| 1137 |
+
"step": 70500
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.9646935739179467,
|
| 1141 |
+
"grad_norm": 1.03636634349823,
|
| 1142 |
+
"learning_rate": 1.8539905975705863e-05,
|
| 1143 |
+
"loss": 1.6863,
|
| 1144 |
+
"mean_token_accuracy": 0.5992070758640766,
|
| 1145 |
+
"step": 71000
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.9714871906356787,
|
| 1149 |
+
"grad_norm": 0.9837082028388977,
|
| 1150 |
+
"learning_rate": 1.5143074540069023e-05,
|
| 1151 |
+
"loss": 1.678,
|
| 1152 |
+
"mean_token_accuracy": 0.5995057925581933,
|
| 1153 |
+
"step": 71500
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"epoch": 0.9782808073534107,
|
| 1157 |
+
"grad_norm": 1.1581671237945557,
|
| 1158 |
+
"learning_rate": 1.1759830430174733e-05,
|
| 1159 |
+
"loss": 1.6811,
|
| 1160 |
+
"mean_token_accuracy": 0.6011576275527477,
|
| 1161 |
+
"step": 72000
|
| 1162 |
+
},
|
| 1163 |
+
{
|
| 1164 |
+
"epoch": 0.9850744240711428,
|
| 1165 |
+
"grad_norm": 0.954434335231781,
|
| 1166 |
+
"learning_rate": 8.362998994537894e-06,
|
| 1167 |
+
"loss": 1.6745,
|
| 1168 |
+
"mean_token_accuracy": 0.6010885013043881,
|
| 1169 |
+
"step": 72500
|
| 1170 |
+
},
|
| 1171 |
+
{
|
| 1172 |
+
"epoch": 0.9918680407888748,
|
| 1173 |
+
"grad_norm": 0.984001100063324,
|
| 1174 |
+
"learning_rate": 4.979754884643605e-06,
|
| 1175 |
+
"loss": 1.6786,
|
| 1176 |
+
"mean_token_accuracy": 0.6019023385941982,
|
| 1177 |
+
"step": 73000
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 0.9986616575066067,
|
| 1181 |
+
"grad_norm": 1.2359673976898193,
|
| 1182 |
+
"learning_rate": 1.5829234490067665e-06,
|
| 1183 |
+
"loss": 1.6652,
|
| 1184 |
+
"mean_token_accuracy": 0.6032795180380345,
|
| 1185 |
+
"step": 73500
|
| 1186 |
+
},
|
| 1187 |
+
{
|
| 1188 |
+
"epoch": 0.9999932063832823,
|
| 1189 |
+
"mean_token_accuracy": 0.5990850239688036,
|
| 1190 |
+
"step": 73598,
|
| 1191 |
+
"total_flos": 9.3961326985583e+17,
|
| 1192 |
+
"train_loss": 1.7179553695028122,
|
| 1193 |
+
"train_runtime": 11634.1539,
|
| 1194 |
+
"train_samples_per_second": 25.304,
|
| 1195 |
+
"train_steps_per_second": 6.326
|
| 1196 |
}
|
| 1197 |
],
|
| 1198 |
"logging_steps": 500,
|
| 1199 |
+
"max_steps": 73598,
|
| 1200 |
"num_input_tokens_seen": 0,
|
| 1201 |
"num_train_epochs": 1,
|
| 1202 |
"save_steps": 0,
|
|
|
|
| 1212 |
"attributes": {}
|
| 1213 |
}
|
| 1214 |
},
|
| 1215 |
+
"total_flos": 9.3961326985583e+17,
|
| 1216 |
"train_batch_size": 1,
|
| 1217 |
"trial_name": null,
|
| 1218 |
"trial_params": null
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 7032
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:48860954eaa763919043f93faa94628bd8ab746ef7fc8bbcc2ca2eb33563b89b
|
| 3 |
size 7032
|