Upload phase3 checkpoint-1476
Browse files- config.json +1 -1
- generation_config.json +1 -1
- model.safetensors +1 -1
- optimizer.pt +1 -1
- rng_state.pth +1 -1
- scheduler.pt +1 -1
- special_tokens_map.json +1 -7
- trainer_state.json +2483 -4
- training_args.bin +1 -1
config.json
CHANGED
|
@@ -24,7 +24,7 @@
|
|
| 24 |
"rope_scaling": null,
|
| 25 |
"rope_theta": 100000.0,
|
| 26 |
"tie_word_embeddings": true,
|
| 27 |
-
"transformers_version": "4.57.
|
| 28 |
"unsloth_version": "2025.12.9",
|
| 29 |
"use_cache": true,
|
| 30 |
"vocab_size": 65537
|
|
|
|
| 24 |
"rope_scaling": null,
|
| 25 |
"rope_theta": 100000.0,
|
| 26 |
"tie_word_embeddings": true,
|
| 27 |
+
"transformers_version": "4.57.1",
|
| 28 |
"unsloth_version": "2025.12.9",
|
| 29 |
"use_cache": true,
|
| 30 |
"vocab_size": 65537
|
generation_config.json
CHANGED
|
@@ -6,5 +6,5 @@
|
|
| 6 |
],
|
| 7 |
"max_length": 8192,
|
| 8 |
"pad_token_id": 2,
|
| 9 |
-
"transformers_version": "4.57.
|
| 10 |
}
|
|
|
|
| 6 |
],
|
| 7 |
"max_length": 8192,
|
| 8 |
"pad_token_id": 2,
|
| 9 |
+
"transformers_version": "4.57.1"
|
| 10 |
}
|
model.safetensors
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1000557344
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:62c4dafbe48c0ed9989a8be4f6e867100bc6208d43171ea8cb23b572df8d41af
|
| 3 |
size 1000557344
|
optimizer.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1101504011
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:e63a549d2219446ec9d9bb6f780ff0fdd3b7af4d8a78b3edd456442d636bbe6c
|
| 3 |
size 1101504011
|
rng_state.pth
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 14645
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:181c5f0270cf39930062ddfa3767a2481d0c360f120b11f8e25dbf533a1cdaba
|
| 3 |
size 14645
|
scheduler.pt
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1465
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:50181b3a62767664c85683bef87af9713c3999da919a398071a03336ecd9c555
|
| 3 |
size 1465
|
special_tokens_map.json
CHANGED
|
@@ -1,12 +1,6 @@
|
|
| 1 |
{
|
| 2 |
"additional_special_tokens": [
|
| 3 |
-
|
| 4 |
-
"content": "<|im_end|>",
|
| 5 |
-
"lstrip": false,
|
| 6 |
-
"normalized": false,
|
| 7 |
-
"rstrip": false,
|
| 8 |
-
"single_word": false
|
| 9 |
-
}
|
| 10 |
],
|
| 11 |
"bos_token": {
|
| 12 |
"content": "<|begin_of_text|>",
|
|
|
|
| 1 |
{
|
| 2 |
"additional_special_tokens": [
|
| 3 |
+
"<|im_end|>"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 4 |
],
|
| 5 |
"bos_token": {
|
| 6 |
"content": "<|begin_of_text|>",
|
trainer_state.json
CHANGED
|
@@ -2,9 +2,9 @@
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
-
"epoch": 0.
|
| 6 |
"eval_steps": 2951,
|
| 7 |
-
"global_step":
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
@@ -18201,6 +18201,2485 @@
|
|
| 18201 |
"learning_rate": 0.02,
|
| 18202 |
"loss": 1.7003,
|
| 18203 |
"step": 2599
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 18204 |
}
|
| 18205 |
],
|
| 18206 |
"logging_steps": 1,
|
|
@@ -18214,13 +20693,13 @@
|
|
| 18214 |
"should_epoch_stop": false,
|
| 18215 |
"should_evaluate": false,
|
| 18216 |
"should_log": false,
|
| 18217 |
-
"should_save":
|
| 18218 |
"should_training_stop": false
|
| 18219 |
},
|
| 18220 |
"attributes": {}
|
| 18221 |
}
|
| 18222 |
},
|
| 18223 |
-
"total_flos": 2.
|
| 18224 |
"train_batch_size": 8,
|
| 18225 |
"trial_name": null,
|
| 18226 |
"trial_params": null
|
|
|
|
| 2 |
"best_global_step": null,
|
| 3 |
"best_metric": null,
|
| 4 |
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.27328588786668984,
|
| 6 |
"eval_steps": 2951,
|
| 7 |
+
"global_step": 2952,
|
| 8 |
"is_hyper_param_search": false,
|
| 9 |
"is_local_process_zero": true,
|
| 10 |
"is_world_process_zero": true,
|
|
|
|
| 18201 |
"learning_rate": 0.02,
|
| 18202 |
"loss": 1.7003,
|
| 18203 |
"step": 2599
|
| 18204 |
+
},
|
| 18205 |
+
{
|
| 18206 |
+
"epoch": 0.24069895272811434,
|
| 18207 |
+
"grad_norm": 0.14453125,
|
| 18208 |
+
"learning_rate": 0.02,
|
| 18209 |
+
"loss": 1.6978,
|
| 18210 |
+
"step": 2600
|
| 18211 |
+
},
|
| 18212 |
+
{
|
| 18213 |
+
"epoch": 0.24079152924839436,
|
| 18214 |
+
"grad_norm": 0.1474609375,
|
| 18215 |
+
"learning_rate": 0.02,
|
| 18216 |
+
"loss": 1.6217,
|
| 18217 |
+
"step": 2601
|
| 18218 |
+
},
|
| 18219 |
+
{
|
| 18220 |
+
"epoch": 0.24088410576867442,
|
| 18221 |
+
"grad_norm": 0.1513671875,
|
| 18222 |
+
"learning_rate": 0.02,
|
| 18223 |
+
"loss": 1.6908,
|
| 18224 |
+
"step": 2602
|
| 18225 |
+
},
|
| 18226 |
+
{
|
| 18227 |
+
"epoch": 0.24097668228895447,
|
| 18228 |
+
"grad_norm": 0.1611328125,
|
| 18229 |
+
"learning_rate": 0.02,
|
| 18230 |
+
"loss": 1.6978,
|
| 18231 |
+
"step": 2603
|
| 18232 |
+
},
|
| 18233 |
+
{
|
| 18234 |
+
"epoch": 0.24106925880923452,
|
| 18235 |
+
"grad_norm": 0.1494140625,
|
| 18236 |
+
"learning_rate": 0.02,
|
| 18237 |
+
"loss": 1.6323,
|
| 18238 |
+
"step": 2604
|
| 18239 |
+
},
|
| 18240 |
+
{
|
| 18241 |
+
"epoch": 0.24116183532951455,
|
| 18242 |
+
"grad_norm": 0.14453125,
|
| 18243 |
+
"learning_rate": 0.02,
|
| 18244 |
+
"loss": 1.7372,
|
| 18245 |
+
"step": 2605
|
| 18246 |
+
},
|
| 18247 |
+
{
|
| 18248 |
+
"epoch": 0.2412544118497946,
|
| 18249 |
+
"grad_norm": 0.1484375,
|
| 18250 |
+
"learning_rate": 0.02,
|
| 18251 |
+
"loss": 1.6353,
|
| 18252 |
+
"step": 2606
|
| 18253 |
+
},
|
| 18254 |
+
{
|
| 18255 |
+
"epoch": 0.24134698837007465,
|
| 18256 |
+
"grad_norm": 0.15234375,
|
| 18257 |
+
"learning_rate": 0.02,
|
| 18258 |
+
"loss": 1.6863,
|
| 18259 |
+
"step": 2607
|
| 18260 |
+
},
|
| 18261 |
+
{
|
| 18262 |
+
"epoch": 0.24143956489035467,
|
| 18263 |
+
"grad_norm": 0.146484375,
|
| 18264 |
+
"learning_rate": 0.02,
|
| 18265 |
+
"loss": 1.6135,
|
| 18266 |
+
"step": 2608
|
| 18267 |
+
},
|
| 18268 |
+
{
|
| 18269 |
+
"epoch": 0.24153214141063473,
|
| 18270 |
+
"grad_norm": 0.154296875,
|
| 18271 |
+
"learning_rate": 0.02,
|
| 18272 |
+
"loss": 1.6775,
|
| 18273 |
+
"step": 2609
|
| 18274 |
+
},
|
| 18275 |
+
{
|
| 18276 |
+
"epoch": 0.24162471793091478,
|
| 18277 |
+
"grad_norm": 0.14453125,
|
| 18278 |
+
"learning_rate": 0.02,
|
| 18279 |
+
"loss": 1.6705,
|
| 18280 |
+
"step": 2610
|
| 18281 |
+
},
|
| 18282 |
+
{
|
| 18283 |
+
"epoch": 0.2417172944511948,
|
| 18284 |
+
"grad_norm": 0.14453125,
|
| 18285 |
+
"learning_rate": 0.02,
|
| 18286 |
+
"loss": 1.6209,
|
| 18287 |
+
"step": 2611
|
| 18288 |
+
},
|
| 18289 |
+
{
|
| 18290 |
+
"epoch": 0.24180987097147486,
|
| 18291 |
+
"grad_norm": 0.134765625,
|
| 18292 |
+
"learning_rate": 0.02,
|
| 18293 |
+
"loss": 1.6261,
|
| 18294 |
+
"step": 2612
|
| 18295 |
+
},
|
| 18296 |
+
{
|
| 18297 |
+
"epoch": 0.2419024474917549,
|
| 18298 |
+
"grad_norm": 0.1376953125,
|
| 18299 |
+
"learning_rate": 0.02,
|
| 18300 |
+
"loss": 1.6419,
|
| 18301 |
+
"step": 2613
|
| 18302 |
+
},
|
| 18303 |
+
{
|
| 18304 |
+
"epoch": 0.24199502401203496,
|
| 18305 |
+
"grad_norm": 0.15234375,
|
| 18306 |
+
"learning_rate": 0.02,
|
| 18307 |
+
"loss": 1.6701,
|
| 18308 |
+
"step": 2614
|
| 18309 |
+
},
|
| 18310 |
+
{
|
| 18311 |
+
"epoch": 0.24208760053231498,
|
| 18312 |
+
"grad_norm": 0.142578125,
|
| 18313 |
+
"learning_rate": 0.02,
|
| 18314 |
+
"loss": 1.687,
|
| 18315 |
+
"step": 2615
|
| 18316 |
+
},
|
| 18317 |
+
{
|
| 18318 |
+
"epoch": 0.24218017705259504,
|
| 18319 |
+
"grad_norm": 0.1435546875,
|
| 18320 |
+
"learning_rate": 0.02,
|
| 18321 |
+
"loss": 1.6425,
|
| 18322 |
+
"step": 2616
|
| 18323 |
+
},
|
| 18324 |
+
{
|
| 18325 |
+
"epoch": 0.2422727535728751,
|
| 18326 |
+
"grad_norm": 0.1484375,
|
| 18327 |
+
"learning_rate": 0.02,
|
| 18328 |
+
"loss": 1.6786,
|
| 18329 |
+
"step": 2617
|
| 18330 |
+
},
|
| 18331 |
+
{
|
| 18332 |
+
"epoch": 0.2423653300931551,
|
| 18333 |
+
"grad_norm": 0.1513671875,
|
| 18334 |
+
"learning_rate": 0.02,
|
| 18335 |
+
"loss": 1.5871,
|
| 18336 |
+
"step": 2618
|
| 18337 |
+
},
|
| 18338 |
+
{
|
| 18339 |
+
"epoch": 0.24245790661343516,
|
| 18340 |
+
"grad_norm": 0.1455078125,
|
| 18341 |
+
"learning_rate": 0.02,
|
| 18342 |
+
"loss": 1.603,
|
| 18343 |
+
"step": 2619
|
| 18344 |
+
},
|
| 18345 |
+
{
|
| 18346 |
+
"epoch": 0.24255048313371522,
|
| 18347 |
+
"grad_norm": 0.15625,
|
| 18348 |
+
"learning_rate": 0.02,
|
| 18349 |
+
"loss": 1.6614,
|
| 18350 |
+
"step": 2620
|
| 18351 |
+
},
|
| 18352 |
+
{
|
| 18353 |
+
"epoch": 0.24264305965399527,
|
| 18354 |
+
"grad_norm": 0.1533203125,
|
| 18355 |
+
"learning_rate": 0.02,
|
| 18356 |
+
"loss": 1.6613,
|
| 18357 |
+
"step": 2621
|
| 18358 |
+
},
|
| 18359 |
+
{
|
| 18360 |
+
"epoch": 0.2427356361742753,
|
| 18361 |
+
"grad_norm": 0.15625,
|
| 18362 |
+
"learning_rate": 0.02,
|
| 18363 |
+
"loss": 1.6533,
|
| 18364 |
+
"step": 2622
|
| 18365 |
+
},
|
| 18366 |
+
{
|
| 18367 |
+
"epoch": 0.24282821269455535,
|
| 18368 |
+
"grad_norm": 0.154296875,
|
| 18369 |
+
"learning_rate": 0.02,
|
| 18370 |
+
"loss": 1.6639,
|
| 18371 |
+
"step": 2623
|
| 18372 |
+
},
|
| 18373 |
+
{
|
| 18374 |
+
"epoch": 0.2429207892148354,
|
| 18375 |
+
"grad_norm": 0.16015625,
|
| 18376 |
+
"learning_rate": 0.02,
|
| 18377 |
+
"loss": 1.6339,
|
| 18378 |
+
"step": 2624
|
| 18379 |
+
},
|
| 18380 |
+
{
|
| 18381 |
+
"epoch": 0.24301336573511542,
|
| 18382 |
+
"grad_norm": 0.154296875,
|
| 18383 |
+
"learning_rate": 0.02,
|
| 18384 |
+
"loss": 1.6194,
|
| 18385 |
+
"step": 2625
|
| 18386 |
+
},
|
| 18387 |
+
{
|
| 18388 |
+
"epoch": 0.24310594225539547,
|
| 18389 |
+
"grad_norm": 0.15234375,
|
| 18390 |
+
"learning_rate": 0.02,
|
| 18391 |
+
"loss": 1.6777,
|
| 18392 |
+
"step": 2626
|
| 18393 |
+
},
|
| 18394 |
+
{
|
| 18395 |
+
"epoch": 0.24319851877567553,
|
| 18396 |
+
"grad_norm": 0.1533203125,
|
| 18397 |
+
"learning_rate": 0.02,
|
| 18398 |
+
"loss": 1.694,
|
| 18399 |
+
"step": 2627
|
| 18400 |
+
},
|
| 18401 |
+
{
|
| 18402 |
+
"epoch": 0.24329109529595555,
|
| 18403 |
+
"grad_norm": 0.14453125,
|
| 18404 |
+
"learning_rate": 0.02,
|
| 18405 |
+
"loss": 1.666,
|
| 18406 |
+
"step": 2628
|
| 18407 |
+
},
|
| 18408 |
+
{
|
| 18409 |
+
"epoch": 0.2433836718162356,
|
| 18410 |
+
"grad_norm": 0.1494140625,
|
| 18411 |
+
"learning_rate": 0.02,
|
| 18412 |
+
"loss": 1.6598,
|
| 18413 |
+
"step": 2629
|
| 18414 |
+
},
|
| 18415 |
+
{
|
| 18416 |
+
"epoch": 0.24347624833651565,
|
| 18417 |
+
"grad_norm": 0.1474609375,
|
| 18418 |
+
"learning_rate": 0.02,
|
| 18419 |
+
"loss": 1.6678,
|
| 18420 |
+
"step": 2630
|
| 18421 |
+
},
|
| 18422 |
+
{
|
| 18423 |
+
"epoch": 0.2435688248567957,
|
| 18424 |
+
"grad_norm": 0.1513671875,
|
| 18425 |
+
"learning_rate": 0.02,
|
| 18426 |
+
"loss": 1.6775,
|
| 18427 |
+
"step": 2631
|
| 18428 |
+
},
|
| 18429 |
+
{
|
| 18430 |
+
"epoch": 0.24366140137707573,
|
| 18431 |
+
"grad_norm": 0.1494140625,
|
| 18432 |
+
"learning_rate": 0.02,
|
| 18433 |
+
"loss": 1.6848,
|
| 18434 |
+
"step": 2632
|
| 18435 |
+
},
|
| 18436 |
+
{
|
| 18437 |
+
"epoch": 0.24375397789735578,
|
| 18438 |
+
"grad_norm": 0.1494140625,
|
| 18439 |
+
"learning_rate": 0.02,
|
| 18440 |
+
"loss": 1.6924,
|
| 18441 |
+
"step": 2633
|
| 18442 |
+
},
|
| 18443 |
+
{
|
| 18444 |
+
"epoch": 0.24384655441763584,
|
| 18445 |
+
"grad_norm": 0.1357421875,
|
| 18446 |
+
"learning_rate": 0.02,
|
| 18447 |
+
"loss": 1.5874,
|
| 18448 |
+
"step": 2634
|
| 18449 |
+
},
|
| 18450 |
+
{
|
| 18451 |
+
"epoch": 0.24393913093791586,
|
| 18452 |
+
"grad_norm": 0.1416015625,
|
| 18453 |
+
"learning_rate": 0.02,
|
| 18454 |
+
"loss": 1.6152,
|
| 18455 |
+
"step": 2635
|
| 18456 |
+
},
|
| 18457 |
+
{
|
| 18458 |
+
"epoch": 0.2440317074581959,
|
| 18459 |
+
"grad_norm": 0.1357421875,
|
| 18460 |
+
"learning_rate": 0.02,
|
| 18461 |
+
"loss": 1.6561,
|
| 18462 |
+
"step": 2636
|
| 18463 |
+
},
|
| 18464 |
+
{
|
| 18465 |
+
"epoch": 0.24412428397847596,
|
| 18466 |
+
"grad_norm": 0.150390625,
|
| 18467 |
+
"learning_rate": 0.02,
|
| 18468 |
+
"loss": 1.6867,
|
| 18469 |
+
"step": 2637
|
| 18470 |
+
},
|
| 18471 |
+
{
|
| 18472 |
+
"epoch": 0.24421686049875602,
|
| 18473 |
+
"grad_norm": 0.16015625,
|
| 18474 |
+
"learning_rate": 0.02,
|
| 18475 |
+
"loss": 1.7139,
|
| 18476 |
+
"step": 2638
|
| 18477 |
+
},
|
| 18478 |
+
{
|
| 18479 |
+
"epoch": 0.24430943701903604,
|
| 18480 |
+
"grad_norm": 0.154296875,
|
| 18481 |
+
"learning_rate": 0.02,
|
| 18482 |
+
"loss": 1.6819,
|
| 18483 |
+
"step": 2639
|
| 18484 |
+
},
|
| 18485 |
+
{
|
| 18486 |
+
"epoch": 0.2444020135393161,
|
| 18487 |
+
"grad_norm": 0.146484375,
|
| 18488 |
+
"learning_rate": 0.02,
|
| 18489 |
+
"loss": 1.6436,
|
| 18490 |
+
"step": 2640
|
| 18491 |
+
},
|
| 18492 |
+
{
|
| 18493 |
+
"epoch": 0.24449459005959615,
|
| 18494 |
+
"grad_norm": 0.1513671875,
|
| 18495 |
+
"learning_rate": 0.02,
|
| 18496 |
+
"loss": 1.6243,
|
| 18497 |
+
"step": 2641
|
| 18498 |
+
},
|
| 18499 |
+
{
|
| 18500 |
+
"epoch": 0.24458716657987617,
|
| 18501 |
+
"grad_norm": 0.1416015625,
|
| 18502 |
+
"learning_rate": 0.02,
|
| 18503 |
+
"loss": 1.6036,
|
| 18504 |
+
"step": 2642
|
| 18505 |
+
},
|
| 18506 |
+
{
|
| 18507 |
+
"epoch": 0.24467974310015622,
|
| 18508 |
+
"grad_norm": 0.150390625,
|
| 18509 |
+
"learning_rate": 0.02,
|
| 18510 |
+
"loss": 1.6353,
|
| 18511 |
+
"step": 2643
|
| 18512 |
+
},
|
| 18513 |
+
{
|
| 18514 |
+
"epoch": 0.24477231962043627,
|
| 18515 |
+
"grad_norm": 0.1494140625,
|
| 18516 |
+
"learning_rate": 0.02,
|
| 18517 |
+
"loss": 1.5888,
|
| 18518 |
+
"step": 2644
|
| 18519 |
+
},
|
| 18520 |
+
{
|
| 18521 |
+
"epoch": 0.2448648961407163,
|
| 18522 |
+
"grad_norm": 0.1455078125,
|
| 18523 |
+
"learning_rate": 0.02,
|
| 18524 |
+
"loss": 1.6343,
|
| 18525 |
+
"step": 2645
|
| 18526 |
+
},
|
| 18527 |
+
{
|
| 18528 |
+
"epoch": 0.24495747266099635,
|
| 18529 |
+
"grad_norm": 0.1396484375,
|
| 18530 |
+
"learning_rate": 0.02,
|
| 18531 |
+
"loss": 1.6279,
|
| 18532 |
+
"step": 2646
|
| 18533 |
+
},
|
| 18534 |
+
{
|
| 18535 |
+
"epoch": 0.2450500491812764,
|
| 18536 |
+
"grad_norm": 0.1435546875,
|
| 18537 |
+
"learning_rate": 0.02,
|
| 18538 |
+
"loss": 1.703,
|
| 18539 |
+
"step": 2647
|
| 18540 |
+
},
|
| 18541 |
+
{
|
| 18542 |
+
"epoch": 0.24514262570155645,
|
| 18543 |
+
"grad_norm": 0.142578125,
|
| 18544 |
+
"learning_rate": 0.02,
|
| 18545 |
+
"loss": 1.6384,
|
| 18546 |
+
"step": 2648
|
| 18547 |
+
},
|
| 18548 |
+
{
|
| 18549 |
+
"epoch": 0.24523520222183648,
|
| 18550 |
+
"grad_norm": 0.1474609375,
|
| 18551 |
+
"learning_rate": 0.02,
|
| 18552 |
+
"loss": 1.6152,
|
| 18553 |
+
"step": 2649
|
| 18554 |
+
},
|
| 18555 |
+
{
|
| 18556 |
+
"epoch": 0.24532777874211653,
|
| 18557 |
+
"grad_norm": 0.150390625,
|
| 18558 |
+
"learning_rate": 0.02,
|
| 18559 |
+
"loss": 1.6887,
|
| 18560 |
+
"step": 2650
|
| 18561 |
+
},
|
| 18562 |
+
{
|
| 18563 |
+
"epoch": 0.24542035526239658,
|
| 18564 |
+
"grad_norm": 0.1494140625,
|
| 18565 |
+
"learning_rate": 0.02,
|
| 18566 |
+
"loss": 1.6758,
|
| 18567 |
+
"step": 2651
|
| 18568 |
+
},
|
| 18569 |
+
{
|
| 18570 |
+
"epoch": 0.2455129317826766,
|
| 18571 |
+
"grad_norm": 0.1494140625,
|
| 18572 |
+
"learning_rate": 0.02,
|
| 18573 |
+
"loss": 1.6623,
|
| 18574 |
+
"step": 2652
|
| 18575 |
+
},
|
| 18576 |
+
{
|
| 18577 |
+
"epoch": 0.24560550830295666,
|
| 18578 |
+
"grad_norm": 0.154296875,
|
| 18579 |
+
"learning_rate": 0.02,
|
| 18580 |
+
"loss": 1.706,
|
| 18581 |
+
"step": 2653
|
| 18582 |
+
},
|
| 18583 |
+
{
|
| 18584 |
+
"epoch": 0.2456980848232367,
|
| 18585 |
+
"grad_norm": 0.1669921875,
|
| 18586 |
+
"learning_rate": 0.02,
|
| 18587 |
+
"loss": 1.7213,
|
| 18588 |
+
"step": 2654
|
| 18589 |
+
},
|
| 18590 |
+
{
|
| 18591 |
+
"epoch": 0.24579066134351676,
|
| 18592 |
+
"grad_norm": 0.1669921875,
|
| 18593 |
+
"learning_rate": 0.02,
|
| 18594 |
+
"loss": 1.7687,
|
| 18595 |
+
"step": 2655
|
| 18596 |
+
},
|
| 18597 |
+
{
|
| 18598 |
+
"epoch": 0.2458832378637968,
|
| 18599 |
+
"grad_norm": 0.1611328125,
|
| 18600 |
+
"learning_rate": 0.02,
|
| 18601 |
+
"loss": 1.7282,
|
| 18602 |
+
"step": 2656
|
| 18603 |
+
},
|
| 18604 |
+
{
|
| 18605 |
+
"epoch": 0.24597581438407684,
|
| 18606 |
+
"grad_norm": 0.1611328125,
|
| 18607 |
+
"learning_rate": 0.02,
|
| 18608 |
+
"loss": 1.7056,
|
| 18609 |
+
"step": 2657
|
| 18610 |
+
},
|
| 18611 |
+
{
|
| 18612 |
+
"epoch": 0.2460683909043569,
|
| 18613 |
+
"grad_norm": 0.15234375,
|
| 18614 |
+
"learning_rate": 0.02,
|
| 18615 |
+
"loss": 1.6426,
|
| 18616 |
+
"step": 2658
|
| 18617 |
+
},
|
| 18618 |
+
{
|
| 18619 |
+
"epoch": 0.24616096742463692,
|
| 18620 |
+
"grad_norm": 0.1552734375,
|
| 18621 |
+
"learning_rate": 0.02,
|
| 18622 |
+
"loss": 1.6115,
|
| 18623 |
+
"step": 2659
|
| 18624 |
+
},
|
| 18625 |
+
{
|
| 18626 |
+
"epoch": 0.24625354394491697,
|
| 18627 |
+
"grad_norm": 0.154296875,
|
| 18628 |
+
"learning_rate": 0.02,
|
| 18629 |
+
"loss": 1.6938,
|
| 18630 |
+
"step": 2660
|
| 18631 |
+
},
|
| 18632 |
+
{
|
| 18633 |
+
"epoch": 0.24634612046519702,
|
| 18634 |
+
"grad_norm": 0.1591796875,
|
| 18635 |
+
"learning_rate": 0.02,
|
| 18636 |
+
"loss": 1.6401,
|
| 18637 |
+
"step": 2661
|
| 18638 |
+
},
|
| 18639 |
+
{
|
| 18640 |
+
"epoch": 0.24643869698547705,
|
| 18641 |
+
"grad_norm": 0.1484375,
|
| 18642 |
+
"learning_rate": 0.02,
|
| 18643 |
+
"loss": 1.6226,
|
| 18644 |
+
"step": 2662
|
| 18645 |
+
},
|
| 18646 |
+
{
|
| 18647 |
+
"epoch": 0.2465312735057571,
|
| 18648 |
+
"grad_norm": 0.158203125,
|
| 18649 |
+
"learning_rate": 0.02,
|
| 18650 |
+
"loss": 1.6953,
|
| 18651 |
+
"step": 2663
|
| 18652 |
+
},
|
| 18653 |
+
{
|
| 18654 |
+
"epoch": 0.24662385002603715,
|
| 18655 |
+
"grad_norm": 0.158203125,
|
| 18656 |
+
"learning_rate": 0.02,
|
| 18657 |
+
"loss": 1.6921,
|
| 18658 |
+
"step": 2664
|
| 18659 |
+
},
|
| 18660 |
+
{
|
| 18661 |
+
"epoch": 0.2467164265463172,
|
| 18662 |
+
"grad_norm": 0.1484375,
|
| 18663 |
+
"learning_rate": 0.02,
|
| 18664 |
+
"loss": 1.6004,
|
| 18665 |
+
"step": 2665
|
| 18666 |
+
},
|
| 18667 |
+
{
|
| 18668 |
+
"epoch": 0.24680900306659723,
|
| 18669 |
+
"grad_norm": 0.1396484375,
|
| 18670 |
+
"learning_rate": 0.02,
|
| 18671 |
+
"loss": 1.6302,
|
| 18672 |
+
"step": 2666
|
| 18673 |
+
},
|
| 18674 |
+
{
|
| 18675 |
+
"epoch": 0.24690157958687728,
|
| 18676 |
+
"grad_norm": 0.1435546875,
|
| 18677 |
+
"learning_rate": 0.02,
|
| 18678 |
+
"loss": 1.604,
|
| 18679 |
+
"step": 2667
|
| 18680 |
+
},
|
| 18681 |
+
{
|
| 18682 |
+
"epoch": 0.24699415610715733,
|
| 18683 |
+
"grad_norm": 0.1494140625,
|
| 18684 |
+
"learning_rate": 0.02,
|
| 18685 |
+
"loss": 1.5945,
|
| 18686 |
+
"step": 2668
|
| 18687 |
+
},
|
| 18688 |
+
{
|
| 18689 |
+
"epoch": 0.24708673262743736,
|
| 18690 |
+
"grad_norm": 0.1630859375,
|
| 18691 |
+
"learning_rate": 0.02,
|
| 18692 |
+
"loss": 1.6713,
|
| 18693 |
+
"step": 2669
|
| 18694 |
+
},
|
| 18695 |
+
{
|
| 18696 |
+
"epoch": 0.2471793091477174,
|
| 18697 |
+
"grad_norm": 0.158203125,
|
| 18698 |
+
"learning_rate": 0.02,
|
| 18699 |
+
"loss": 1.6453,
|
| 18700 |
+
"step": 2670
|
| 18701 |
+
},
|
| 18702 |
+
{
|
| 18703 |
+
"epoch": 0.24727188566799746,
|
| 18704 |
+
"grad_norm": 0.15234375,
|
| 18705 |
+
"learning_rate": 0.02,
|
| 18706 |
+
"loss": 1.6212,
|
| 18707 |
+
"step": 2671
|
| 18708 |
+
},
|
| 18709 |
+
{
|
| 18710 |
+
"epoch": 0.2473644621882775,
|
| 18711 |
+
"grad_norm": 0.1552734375,
|
| 18712 |
+
"learning_rate": 0.02,
|
| 18713 |
+
"loss": 1.6691,
|
| 18714 |
+
"step": 2672
|
| 18715 |
+
},
|
| 18716 |
+
{
|
| 18717 |
+
"epoch": 0.24745703870855754,
|
| 18718 |
+
"grad_norm": 0.16015625,
|
| 18719 |
+
"learning_rate": 0.02,
|
| 18720 |
+
"loss": 1.6895,
|
| 18721 |
+
"step": 2673
|
| 18722 |
+
},
|
| 18723 |
+
{
|
| 18724 |
+
"epoch": 0.2475496152288376,
|
| 18725 |
+
"grad_norm": 0.1494140625,
|
| 18726 |
+
"learning_rate": 0.02,
|
| 18727 |
+
"loss": 1.6548,
|
| 18728 |
+
"step": 2674
|
| 18729 |
+
},
|
| 18730 |
+
{
|
| 18731 |
+
"epoch": 0.24764219174911764,
|
| 18732 |
+
"grad_norm": 0.15625,
|
| 18733 |
+
"learning_rate": 0.02,
|
| 18734 |
+
"loss": 1.6663,
|
| 18735 |
+
"step": 2675
|
| 18736 |
+
},
|
| 18737 |
+
{
|
| 18738 |
+
"epoch": 0.24773476826939766,
|
| 18739 |
+
"grad_norm": 0.15234375,
|
| 18740 |
+
"learning_rate": 0.02,
|
| 18741 |
+
"loss": 1.6238,
|
| 18742 |
+
"step": 2676
|
| 18743 |
+
},
|
| 18744 |
+
{
|
| 18745 |
+
"epoch": 0.24782734478967772,
|
| 18746 |
+
"grad_norm": 0.16015625,
|
| 18747 |
+
"learning_rate": 0.02,
|
| 18748 |
+
"loss": 1.6859,
|
| 18749 |
+
"step": 2677
|
| 18750 |
+
},
|
| 18751 |
+
{
|
| 18752 |
+
"epoch": 0.24791992130995777,
|
| 18753 |
+
"grad_norm": 0.1630859375,
|
| 18754 |
+
"learning_rate": 0.02,
|
| 18755 |
+
"loss": 1.5579,
|
| 18756 |
+
"step": 2678
|
| 18757 |
+
},
|
| 18758 |
+
{
|
| 18759 |
+
"epoch": 0.2480124978302378,
|
| 18760 |
+
"grad_norm": 0.1572265625,
|
| 18761 |
+
"learning_rate": 0.02,
|
| 18762 |
+
"loss": 1.6192,
|
| 18763 |
+
"step": 2679
|
| 18764 |
+
},
|
| 18765 |
+
{
|
| 18766 |
+
"epoch": 0.24810507435051785,
|
| 18767 |
+
"grad_norm": 0.1552734375,
|
| 18768 |
+
"learning_rate": 0.02,
|
| 18769 |
+
"loss": 1.6615,
|
| 18770 |
+
"step": 2680
|
| 18771 |
+
},
|
| 18772 |
+
{
|
| 18773 |
+
"epoch": 0.2481976508707979,
|
| 18774 |
+
"grad_norm": 0.1552734375,
|
| 18775 |
+
"learning_rate": 0.02,
|
| 18776 |
+
"loss": 1.6472,
|
| 18777 |
+
"step": 2681
|
| 18778 |
+
},
|
| 18779 |
+
{
|
| 18780 |
+
"epoch": 0.24829022739107795,
|
| 18781 |
+
"grad_norm": 0.1484375,
|
| 18782 |
+
"learning_rate": 0.02,
|
| 18783 |
+
"loss": 1.6743,
|
| 18784 |
+
"step": 2682
|
| 18785 |
+
},
|
| 18786 |
+
{
|
| 18787 |
+
"epoch": 0.24838280391135797,
|
| 18788 |
+
"grad_norm": 0.150390625,
|
| 18789 |
+
"learning_rate": 0.02,
|
| 18790 |
+
"loss": 1.6743,
|
| 18791 |
+
"step": 2683
|
| 18792 |
+
},
|
| 18793 |
+
{
|
| 18794 |
+
"epoch": 0.24847538043163803,
|
| 18795 |
+
"grad_norm": 0.140625,
|
| 18796 |
+
"learning_rate": 0.02,
|
| 18797 |
+
"loss": 1.676,
|
| 18798 |
+
"step": 2684
|
| 18799 |
+
},
|
| 18800 |
+
{
|
| 18801 |
+
"epoch": 0.24856795695191808,
|
| 18802 |
+
"grad_norm": 0.1396484375,
|
| 18803 |
+
"learning_rate": 0.02,
|
| 18804 |
+
"loss": 1.6005,
|
| 18805 |
+
"step": 2685
|
| 18806 |
+
},
|
| 18807 |
+
{
|
| 18808 |
+
"epoch": 0.2486605334721981,
|
| 18809 |
+
"grad_norm": 0.13671875,
|
| 18810 |
+
"learning_rate": 0.02,
|
| 18811 |
+
"loss": 1.6041,
|
| 18812 |
+
"step": 2686
|
| 18813 |
+
},
|
| 18814 |
+
{
|
| 18815 |
+
"epoch": 0.24875310999247816,
|
| 18816 |
+
"grad_norm": 0.1455078125,
|
| 18817 |
+
"learning_rate": 0.02,
|
| 18818 |
+
"loss": 1.6173,
|
| 18819 |
+
"step": 2687
|
| 18820 |
+
},
|
| 18821 |
+
{
|
| 18822 |
+
"epoch": 0.2488456865127582,
|
| 18823 |
+
"grad_norm": 0.13671875,
|
| 18824 |
+
"learning_rate": 0.02,
|
| 18825 |
+
"loss": 1.6514,
|
| 18826 |
+
"step": 2688
|
| 18827 |
+
},
|
| 18828 |
+
{
|
| 18829 |
+
"epoch": 0.24893826303303826,
|
| 18830 |
+
"grad_norm": 0.1376953125,
|
| 18831 |
+
"learning_rate": 0.02,
|
| 18832 |
+
"loss": 1.6773,
|
| 18833 |
+
"step": 2689
|
| 18834 |
+
},
|
| 18835 |
+
{
|
| 18836 |
+
"epoch": 0.24903083955331828,
|
| 18837 |
+
"grad_norm": 0.14453125,
|
| 18838 |
+
"learning_rate": 0.02,
|
| 18839 |
+
"loss": 1.6496,
|
| 18840 |
+
"step": 2690
|
| 18841 |
+
},
|
| 18842 |
+
{
|
| 18843 |
+
"epoch": 0.24912341607359834,
|
| 18844 |
+
"grad_norm": 0.1572265625,
|
| 18845 |
+
"learning_rate": 0.02,
|
| 18846 |
+
"loss": 1.6966,
|
| 18847 |
+
"step": 2691
|
| 18848 |
+
},
|
| 18849 |
+
{
|
| 18850 |
+
"epoch": 0.2492159925938784,
|
| 18851 |
+
"grad_norm": 0.1513671875,
|
| 18852 |
+
"learning_rate": 0.02,
|
| 18853 |
+
"loss": 1.6741,
|
| 18854 |
+
"step": 2692
|
| 18855 |
+
},
|
| 18856 |
+
{
|
| 18857 |
+
"epoch": 0.2493085691141584,
|
| 18858 |
+
"grad_norm": 0.166015625,
|
| 18859 |
+
"learning_rate": 0.02,
|
| 18860 |
+
"loss": 1.6864,
|
| 18861 |
+
"step": 2693
|
| 18862 |
+
},
|
| 18863 |
+
{
|
| 18864 |
+
"epoch": 0.24940114563443846,
|
| 18865 |
+
"grad_norm": 0.1572265625,
|
| 18866 |
+
"learning_rate": 0.02,
|
| 18867 |
+
"loss": 1.6526,
|
| 18868 |
+
"step": 2694
|
| 18869 |
+
},
|
| 18870 |
+
{
|
| 18871 |
+
"epoch": 0.24949372215471852,
|
| 18872 |
+
"grad_norm": 0.1611328125,
|
| 18873 |
+
"learning_rate": 0.02,
|
| 18874 |
+
"loss": 1.7017,
|
| 18875 |
+
"step": 2695
|
| 18876 |
+
},
|
| 18877 |
+
{
|
| 18878 |
+
"epoch": 0.24958629867499854,
|
| 18879 |
+
"grad_norm": 0.1533203125,
|
| 18880 |
+
"learning_rate": 0.02,
|
| 18881 |
+
"loss": 1.6501,
|
| 18882 |
+
"step": 2696
|
| 18883 |
+
},
|
| 18884 |
+
{
|
| 18885 |
+
"epoch": 0.2496788751952786,
|
| 18886 |
+
"grad_norm": 0.16015625,
|
| 18887 |
+
"learning_rate": 0.02,
|
| 18888 |
+
"loss": 1.6995,
|
| 18889 |
+
"step": 2697
|
| 18890 |
+
},
|
| 18891 |
+
{
|
| 18892 |
+
"epoch": 0.24977145171555865,
|
| 18893 |
+
"grad_norm": 0.1552734375,
|
| 18894 |
+
"learning_rate": 0.02,
|
| 18895 |
+
"loss": 1.6778,
|
| 18896 |
+
"step": 2698
|
| 18897 |
+
},
|
| 18898 |
+
{
|
| 18899 |
+
"epoch": 0.2498640282358387,
|
| 18900 |
+
"grad_norm": 0.1552734375,
|
| 18901 |
+
"learning_rate": 0.02,
|
| 18902 |
+
"loss": 1.6304,
|
| 18903 |
+
"step": 2699
|
| 18904 |
+
},
|
| 18905 |
+
{
|
| 18906 |
+
"epoch": 0.24995660475611872,
|
| 18907 |
+
"grad_norm": 0.142578125,
|
| 18908 |
+
"learning_rate": 0.02,
|
| 18909 |
+
"loss": 1.7184,
|
| 18910 |
+
"step": 2700
|
| 18911 |
+
},
|
| 18912 |
+
{
|
| 18913 |
+
"epoch": 0.25004918127639875,
|
| 18914 |
+
"grad_norm": 0.16015625,
|
| 18915 |
+
"learning_rate": 0.02,
|
| 18916 |
+
"loss": 1.7369,
|
| 18917 |
+
"step": 2701
|
| 18918 |
+
},
|
| 18919 |
+
{
|
| 18920 |
+
"epoch": 0.2501417577966788,
|
| 18921 |
+
"grad_norm": 0.1484375,
|
| 18922 |
+
"learning_rate": 0.02,
|
| 18923 |
+
"loss": 1.6094,
|
| 18924 |
+
"step": 2702
|
| 18925 |
+
},
|
| 18926 |
+
{
|
| 18927 |
+
"epoch": 0.25023433431695885,
|
| 18928 |
+
"grad_norm": 0.1533203125,
|
| 18929 |
+
"learning_rate": 0.02,
|
| 18930 |
+
"loss": 1.6358,
|
| 18931 |
+
"step": 2703
|
| 18932 |
+
},
|
| 18933 |
+
{
|
| 18934 |
+
"epoch": 0.2503269108372389,
|
| 18935 |
+
"grad_norm": 0.15234375,
|
| 18936 |
+
"learning_rate": 0.02,
|
| 18937 |
+
"loss": 1.6927,
|
| 18938 |
+
"step": 2704
|
| 18939 |
+
},
|
| 18940 |
+
{
|
| 18941 |
+
"epoch": 0.25041948735751896,
|
| 18942 |
+
"grad_norm": 0.1572265625,
|
| 18943 |
+
"learning_rate": 0.02,
|
| 18944 |
+
"loss": 1.6453,
|
| 18945 |
+
"step": 2705
|
| 18946 |
+
},
|
| 18947 |
+
{
|
| 18948 |
+
"epoch": 0.250512063877799,
|
| 18949 |
+
"grad_norm": 0.1728515625,
|
| 18950 |
+
"learning_rate": 0.02,
|
| 18951 |
+
"loss": 1.6397,
|
| 18952 |
+
"step": 2706
|
| 18953 |
+
},
|
| 18954 |
+
{
|
| 18955 |
+
"epoch": 0.25060464039807906,
|
| 18956 |
+
"grad_norm": 0.1533203125,
|
| 18957 |
+
"learning_rate": 0.02,
|
| 18958 |
+
"loss": 1.6373,
|
| 18959 |
+
"step": 2707
|
| 18960 |
+
},
|
| 18961 |
+
{
|
| 18962 |
+
"epoch": 0.25069721691835906,
|
| 18963 |
+
"grad_norm": 0.1572265625,
|
| 18964 |
+
"learning_rate": 0.02,
|
| 18965 |
+
"loss": 1.6574,
|
| 18966 |
+
"step": 2708
|
| 18967 |
+
},
|
| 18968 |
+
{
|
| 18969 |
+
"epoch": 0.2507897934386391,
|
| 18970 |
+
"grad_norm": 0.1669921875,
|
| 18971 |
+
"learning_rate": 0.02,
|
| 18972 |
+
"loss": 1.683,
|
| 18973 |
+
"step": 2709
|
| 18974 |
+
},
|
| 18975 |
+
{
|
| 18976 |
+
"epoch": 0.25088236995891916,
|
| 18977 |
+
"grad_norm": 0.1640625,
|
| 18978 |
+
"learning_rate": 0.02,
|
| 18979 |
+
"loss": 1.7074,
|
| 18980 |
+
"step": 2710
|
| 18981 |
+
},
|
| 18982 |
+
{
|
| 18983 |
+
"epoch": 0.2509749464791992,
|
| 18984 |
+
"grad_norm": 0.158203125,
|
| 18985 |
+
"learning_rate": 0.02,
|
| 18986 |
+
"loss": 1.6745,
|
| 18987 |
+
"step": 2711
|
| 18988 |
+
},
|
| 18989 |
+
{
|
| 18990 |
+
"epoch": 0.25106752299947926,
|
| 18991 |
+
"grad_norm": 0.1552734375,
|
| 18992 |
+
"learning_rate": 0.02,
|
| 18993 |
+
"loss": 1.6705,
|
| 18994 |
+
"step": 2712
|
| 18995 |
+
},
|
| 18996 |
+
{
|
| 18997 |
+
"epoch": 0.2511600995197593,
|
| 18998 |
+
"grad_norm": 0.162109375,
|
| 18999 |
+
"learning_rate": 0.02,
|
| 19000 |
+
"loss": 1.6188,
|
| 19001 |
+
"step": 2713
|
| 19002 |
+
},
|
| 19003 |
+
{
|
| 19004 |
+
"epoch": 0.25125267604003937,
|
| 19005 |
+
"grad_norm": 0.1474609375,
|
| 19006 |
+
"learning_rate": 0.02,
|
| 19007 |
+
"loss": 1.6323,
|
| 19008 |
+
"step": 2714
|
| 19009 |
+
},
|
| 19010 |
+
{
|
| 19011 |
+
"epoch": 0.25134525256031937,
|
| 19012 |
+
"grad_norm": 0.158203125,
|
| 19013 |
+
"learning_rate": 0.02,
|
| 19014 |
+
"loss": 1.6792,
|
| 19015 |
+
"step": 2715
|
| 19016 |
+
},
|
| 19017 |
+
{
|
| 19018 |
+
"epoch": 0.2514378290805994,
|
| 19019 |
+
"grad_norm": 0.1552734375,
|
| 19020 |
+
"learning_rate": 0.02,
|
| 19021 |
+
"loss": 1.7093,
|
| 19022 |
+
"step": 2716
|
| 19023 |
+
},
|
| 19024 |
+
{
|
| 19025 |
+
"epoch": 0.25153040560087947,
|
| 19026 |
+
"grad_norm": 0.158203125,
|
| 19027 |
+
"learning_rate": 0.02,
|
| 19028 |
+
"loss": 1.6941,
|
| 19029 |
+
"step": 2717
|
| 19030 |
+
},
|
| 19031 |
+
{
|
| 19032 |
+
"epoch": 0.2516229821211595,
|
| 19033 |
+
"grad_norm": 0.1474609375,
|
| 19034 |
+
"learning_rate": 0.02,
|
| 19035 |
+
"loss": 1.6991,
|
| 19036 |
+
"step": 2718
|
| 19037 |
+
},
|
| 19038 |
+
{
|
| 19039 |
+
"epoch": 0.2517155586414396,
|
| 19040 |
+
"grad_norm": 0.150390625,
|
| 19041 |
+
"learning_rate": 0.02,
|
| 19042 |
+
"loss": 1.6604,
|
| 19043 |
+
"step": 2719
|
| 19044 |
+
},
|
| 19045 |
+
{
|
| 19046 |
+
"epoch": 0.2518081351617196,
|
| 19047 |
+
"grad_norm": 0.142578125,
|
| 19048 |
+
"learning_rate": 0.02,
|
| 19049 |
+
"loss": 1.7023,
|
| 19050 |
+
"step": 2720
|
| 19051 |
+
},
|
| 19052 |
+
{
|
| 19053 |
+
"epoch": 0.2519007116819997,
|
| 19054 |
+
"grad_norm": 0.1552734375,
|
| 19055 |
+
"learning_rate": 0.02,
|
| 19056 |
+
"loss": 1.72,
|
| 19057 |
+
"step": 2721
|
| 19058 |
+
},
|
| 19059 |
+
{
|
| 19060 |
+
"epoch": 0.2519932882022797,
|
| 19061 |
+
"grad_norm": 0.1474609375,
|
| 19062 |
+
"learning_rate": 0.02,
|
| 19063 |
+
"loss": 1.6754,
|
| 19064 |
+
"step": 2722
|
| 19065 |
+
},
|
| 19066 |
+
{
|
| 19067 |
+
"epoch": 0.2520858647225597,
|
| 19068 |
+
"grad_norm": 0.146484375,
|
| 19069 |
+
"learning_rate": 0.02,
|
| 19070 |
+
"loss": 1.713,
|
| 19071 |
+
"step": 2723
|
| 19072 |
+
},
|
| 19073 |
+
{
|
| 19074 |
+
"epoch": 0.2521784412428398,
|
| 19075 |
+
"grad_norm": 0.150390625,
|
| 19076 |
+
"learning_rate": 0.02,
|
| 19077 |
+
"loss": 1.728,
|
| 19078 |
+
"step": 2724
|
| 19079 |
+
},
|
| 19080 |
+
{
|
| 19081 |
+
"epoch": 0.25227101776311983,
|
| 19082 |
+
"grad_norm": 0.1533203125,
|
| 19083 |
+
"learning_rate": 0.02,
|
| 19084 |
+
"loss": 1.6826,
|
| 19085 |
+
"step": 2725
|
| 19086 |
+
},
|
| 19087 |
+
{
|
| 19088 |
+
"epoch": 0.2523635942833999,
|
| 19089 |
+
"grad_norm": 0.158203125,
|
| 19090 |
+
"learning_rate": 0.02,
|
| 19091 |
+
"loss": 1.6465,
|
| 19092 |
+
"step": 2726
|
| 19093 |
+
},
|
| 19094 |
+
{
|
| 19095 |
+
"epoch": 0.25245617080367994,
|
| 19096 |
+
"grad_norm": 0.1533203125,
|
| 19097 |
+
"learning_rate": 0.02,
|
| 19098 |
+
"loss": 1.6399,
|
| 19099 |
+
"step": 2727
|
| 19100 |
+
},
|
| 19101 |
+
{
|
| 19102 |
+
"epoch": 0.25254874732396,
|
| 19103 |
+
"grad_norm": 0.1435546875,
|
| 19104 |
+
"learning_rate": 0.02,
|
| 19105 |
+
"loss": 1.6052,
|
| 19106 |
+
"step": 2728
|
| 19107 |
+
},
|
| 19108 |
+
{
|
| 19109 |
+
"epoch": 0.25264132384424,
|
| 19110 |
+
"grad_norm": 0.1416015625,
|
| 19111 |
+
"learning_rate": 0.02,
|
| 19112 |
+
"loss": 1.5989,
|
| 19113 |
+
"step": 2729
|
| 19114 |
+
},
|
| 19115 |
+
{
|
| 19116 |
+
"epoch": 0.25273390036452004,
|
| 19117 |
+
"grad_norm": 0.14453125,
|
| 19118 |
+
"learning_rate": 0.02,
|
| 19119 |
+
"loss": 1.6454,
|
| 19120 |
+
"step": 2730
|
| 19121 |
+
},
|
| 19122 |
+
{
|
| 19123 |
+
"epoch": 0.2528264768848001,
|
| 19124 |
+
"grad_norm": 0.1533203125,
|
| 19125 |
+
"learning_rate": 0.02,
|
| 19126 |
+
"loss": 1.6814,
|
| 19127 |
+
"step": 2731
|
| 19128 |
+
},
|
| 19129 |
+
{
|
| 19130 |
+
"epoch": 0.25291905340508014,
|
| 19131 |
+
"grad_norm": 0.166015625,
|
| 19132 |
+
"learning_rate": 0.02,
|
| 19133 |
+
"loss": 1.6407,
|
| 19134 |
+
"step": 2732
|
| 19135 |
+
},
|
| 19136 |
+
{
|
| 19137 |
+
"epoch": 0.2530116299253602,
|
| 19138 |
+
"grad_norm": 0.1484375,
|
| 19139 |
+
"learning_rate": 0.02,
|
| 19140 |
+
"loss": 1.6849,
|
| 19141 |
+
"step": 2733
|
| 19142 |
+
},
|
| 19143 |
+
{
|
| 19144 |
+
"epoch": 0.25310420644564025,
|
| 19145 |
+
"grad_norm": 0.1552734375,
|
| 19146 |
+
"learning_rate": 0.02,
|
| 19147 |
+
"loss": 1.5967,
|
| 19148 |
+
"step": 2734
|
| 19149 |
+
},
|
| 19150 |
+
{
|
| 19151 |
+
"epoch": 0.25319678296592024,
|
| 19152 |
+
"grad_norm": 0.1640625,
|
| 19153 |
+
"learning_rate": 0.02,
|
| 19154 |
+
"loss": 1.6596,
|
| 19155 |
+
"step": 2735
|
| 19156 |
+
},
|
| 19157 |
+
{
|
| 19158 |
+
"epoch": 0.2532893594862003,
|
| 19159 |
+
"grad_norm": 0.15234375,
|
| 19160 |
+
"learning_rate": 0.02,
|
| 19161 |
+
"loss": 1.6836,
|
| 19162 |
+
"step": 2736
|
| 19163 |
+
},
|
| 19164 |
+
{
|
| 19165 |
+
"epoch": 0.25338193600648035,
|
| 19166 |
+
"grad_norm": 0.146484375,
|
| 19167 |
+
"learning_rate": 0.02,
|
| 19168 |
+
"loss": 1.689,
|
| 19169 |
+
"step": 2737
|
| 19170 |
+
},
|
| 19171 |
+
{
|
| 19172 |
+
"epoch": 0.2534745125267604,
|
| 19173 |
+
"grad_norm": 0.150390625,
|
| 19174 |
+
"learning_rate": 0.02,
|
| 19175 |
+
"loss": 1.7011,
|
| 19176 |
+
"step": 2738
|
| 19177 |
+
},
|
| 19178 |
+
{
|
| 19179 |
+
"epoch": 0.25356708904704045,
|
| 19180 |
+
"grad_norm": 0.140625,
|
| 19181 |
+
"learning_rate": 0.02,
|
| 19182 |
+
"loss": 1.6348,
|
| 19183 |
+
"step": 2739
|
| 19184 |
+
},
|
| 19185 |
+
{
|
| 19186 |
+
"epoch": 0.2536596655673205,
|
| 19187 |
+
"grad_norm": 0.142578125,
|
| 19188 |
+
"learning_rate": 0.02,
|
| 19189 |
+
"loss": 1.6331,
|
| 19190 |
+
"step": 2740
|
| 19191 |
+
},
|
| 19192 |
+
{
|
| 19193 |
+
"epoch": 0.25375224208760055,
|
| 19194 |
+
"grad_norm": 0.1484375,
|
| 19195 |
+
"learning_rate": 0.02,
|
| 19196 |
+
"loss": 1.62,
|
| 19197 |
+
"step": 2741
|
| 19198 |
+
},
|
| 19199 |
+
{
|
| 19200 |
+
"epoch": 0.25384481860788055,
|
| 19201 |
+
"grad_norm": 0.146484375,
|
| 19202 |
+
"learning_rate": 0.02,
|
| 19203 |
+
"loss": 1.6445,
|
| 19204 |
+
"step": 2742
|
| 19205 |
+
},
|
| 19206 |
+
{
|
| 19207 |
+
"epoch": 0.2539373951281606,
|
| 19208 |
+
"grad_norm": 0.1416015625,
|
| 19209 |
+
"learning_rate": 0.02,
|
| 19210 |
+
"loss": 1.6724,
|
| 19211 |
+
"step": 2743
|
| 19212 |
+
},
|
| 19213 |
+
{
|
| 19214 |
+
"epoch": 0.25402997164844066,
|
| 19215 |
+
"grad_norm": 0.150390625,
|
| 19216 |
+
"learning_rate": 0.02,
|
| 19217 |
+
"loss": 1.6466,
|
| 19218 |
+
"step": 2744
|
| 19219 |
+
},
|
| 19220 |
+
{
|
| 19221 |
+
"epoch": 0.2541225481687207,
|
| 19222 |
+
"grad_norm": 0.146484375,
|
| 19223 |
+
"learning_rate": 0.02,
|
| 19224 |
+
"loss": 1.689,
|
| 19225 |
+
"step": 2745
|
| 19226 |
+
},
|
| 19227 |
+
{
|
| 19228 |
+
"epoch": 0.25421512468900076,
|
| 19229 |
+
"grad_norm": 0.1494140625,
|
| 19230 |
+
"learning_rate": 0.02,
|
| 19231 |
+
"loss": 1.6311,
|
| 19232 |
+
"step": 2746
|
| 19233 |
+
},
|
| 19234 |
+
{
|
| 19235 |
+
"epoch": 0.2543077012092808,
|
| 19236 |
+
"grad_norm": 0.158203125,
|
| 19237 |
+
"learning_rate": 0.02,
|
| 19238 |
+
"loss": 1.6765,
|
| 19239 |
+
"step": 2747
|
| 19240 |
+
},
|
| 19241 |
+
{
|
| 19242 |
+
"epoch": 0.25440027772956086,
|
| 19243 |
+
"grad_norm": 0.1611328125,
|
| 19244 |
+
"learning_rate": 0.02,
|
| 19245 |
+
"loss": 1.6515,
|
| 19246 |
+
"step": 2748
|
| 19247 |
+
},
|
| 19248 |
+
{
|
| 19249 |
+
"epoch": 0.25449285424984086,
|
| 19250 |
+
"grad_norm": 0.15234375,
|
| 19251 |
+
"learning_rate": 0.02,
|
| 19252 |
+
"loss": 1.6626,
|
| 19253 |
+
"step": 2749
|
| 19254 |
+
},
|
| 19255 |
+
{
|
| 19256 |
+
"epoch": 0.2545854307701209,
|
| 19257 |
+
"grad_norm": 0.158203125,
|
| 19258 |
+
"learning_rate": 0.02,
|
| 19259 |
+
"loss": 1.7178,
|
| 19260 |
+
"step": 2750
|
| 19261 |
+
},
|
| 19262 |
+
{
|
| 19263 |
+
"epoch": 0.25467800729040097,
|
| 19264 |
+
"grad_norm": 0.1513671875,
|
| 19265 |
+
"learning_rate": 0.02,
|
| 19266 |
+
"loss": 1.6839,
|
| 19267 |
+
"step": 2751
|
| 19268 |
+
},
|
| 19269 |
+
{
|
| 19270 |
+
"epoch": 0.254770583810681,
|
| 19271 |
+
"grad_norm": 0.1572265625,
|
| 19272 |
+
"learning_rate": 0.02,
|
| 19273 |
+
"loss": 1.6863,
|
| 19274 |
+
"step": 2752
|
| 19275 |
+
},
|
| 19276 |
+
{
|
| 19277 |
+
"epoch": 0.25486316033096107,
|
| 19278 |
+
"grad_norm": 0.1552734375,
|
| 19279 |
+
"learning_rate": 0.02,
|
| 19280 |
+
"loss": 1.6898,
|
| 19281 |
+
"step": 2753
|
| 19282 |
+
},
|
| 19283 |
+
{
|
| 19284 |
+
"epoch": 0.2549557368512411,
|
| 19285 |
+
"grad_norm": 0.1630859375,
|
| 19286 |
+
"learning_rate": 0.02,
|
| 19287 |
+
"loss": 1.609,
|
| 19288 |
+
"step": 2754
|
| 19289 |
+
},
|
| 19290 |
+
{
|
| 19291 |
+
"epoch": 0.2550483133715212,
|
| 19292 |
+
"grad_norm": 0.1533203125,
|
| 19293 |
+
"learning_rate": 0.02,
|
| 19294 |
+
"loss": 1.616,
|
| 19295 |
+
"step": 2755
|
| 19296 |
+
},
|
| 19297 |
+
{
|
| 19298 |
+
"epoch": 0.25514088989180117,
|
| 19299 |
+
"grad_norm": 0.158203125,
|
| 19300 |
+
"learning_rate": 0.02,
|
| 19301 |
+
"loss": 1.6102,
|
| 19302 |
+
"step": 2756
|
| 19303 |
+
},
|
| 19304 |
+
{
|
| 19305 |
+
"epoch": 0.2552334664120812,
|
| 19306 |
+
"grad_norm": 0.1572265625,
|
| 19307 |
+
"learning_rate": 0.02,
|
| 19308 |
+
"loss": 1.6667,
|
| 19309 |
+
"step": 2757
|
| 19310 |
+
},
|
| 19311 |
+
{
|
| 19312 |
+
"epoch": 0.2553260429323613,
|
| 19313 |
+
"grad_norm": 0.154296875,
|
| 19314 |
+
"learning_rate": 0.02,
|
| 19315 |
+
"loss": 1.7157,
|
| 19316 |
+
"step": 2758
|
| 19317 |
+
},
|
| 19318 |
+
{
|
| 19319 |
+
"epoch": 0.2554186194526413,
|
| 19320 |
+
"grad_norm": 0.150390625,
|
| 19321 |
+
"learning_rate": 0.02,
|
| 19322 |
+
"loss": 1.6555,
|
| 19323 |
+
"step": 2759
|
| 19324 |
+
},
|
| 19325 |
+
{
|
| 19326 |
+
"epoch": 0.2555111959729214,
|
| 19327 |
+
"grad_norm": 0.142578125,
|
| 19328 |
+
"learning_rate": 0.02,
|
| 19329 |
+
"loss": 1.6451,
|
| 19330 |
+
"step": 2760
|
| 19331 |
+
},
|
| 19332 |
+
{
|
| 19333 |
+
"epoch": 0.25560377249320143,
|
| 19334 |
+
"grad_norm": 0.154296875,
|
| 19335 |
+
"learning_rate": 0.02,
|
| 19336 |
+
"loss": 1.5886,
|
| 19337 |
+
"step": 2761
|
| 19338 |
+
},
|
| 19339 |
+
{
|
| 19340 |
+
"epoch": 0.2556963490134815,
|
| 19341 |
+
"grad_norm": 0.1416015625,
|
| 19342 |
+
"learning_rate": 0.02,
|
| 19343 |
+
"loss": 1.6771,
|
| 19344 |
+
"step": 2762
|
| 19345 |
+
},
|
| 19346 |
+
{
|
| 19347 |
+
"epoch": 0.2557889255337615,
|
| 19348 |
+
"grad_norm": 0.1572265625,
|
| 19349 |
+
"learning_rate": 0.02,
|
| 19350 |
+
"loss": 1.6772,
|
| 19351 |
+
"step": 2763
|
| 19352 |
+
},
|
| 19353 |
+
{
|
| 19354 |
+
"epoch": 0.25588150205404153,
|
| 19355 |
+
"grad_norm": 0.1591796875,
|
| 19356 |
+
"learning_rate": 0.02,
|
| 19357 |
+
"loss": 1.6507,
|
| 19358 |
+
"step": 2764
|
| 19359 |
+
},
|
| 19360 |
+
{
|
| 19361 |
+
"epoch": 0.2559740785743216,
|
| 19362 |
+
"grad_norm": 0.15234375,
|
| 19363 |
+
"learning_rate": 0.02,
|
| 19364 |
+
"loss": 1.6707,
|
| 19365 |
+
"step": 2765
|
| 19366 |
+
},
|
| 19367 |
+
{
|
| 19368 |
+
"epoch": 0.25606665509460164,
|
| 19369 |
+
"grad_norm": 0.14453125,
|
| 19370 |
+
"learning_rate": 0.02,
|
| 19371 |
+
"loss": 1.624,
|
| 19372 |
+
"step": 2766
|
| 19373 |
+
},
|
| 19374 |
+
{
|
| 19375 |
+
"epoch": 0.2561592316148817,
|
| 19376 |
+
"grad_norm": 0.140625,
|
| 19377 |
+
"learning_rate": 0.02,
|
| 19378 |
+
"loss": 1.6516,
|
| 19379 |
+
"step": 2767
|
| 19380 |
+
},
|
| 19381 |
+
{
|
| 19382 |
+
"epoch": 0.25625180813516174,
|
| 19383 |
+
"grad_norm": 0.1513671875,
|
| 19384 |
+
"learning_rate": 0.02,
|
| 19385 |
+
"loss": 1.6816,
|
| 19386 |
+
"step": 2768
|
| 19387 |
+
},
|
| 19388 |
+
{
|
| 19389 |
+
"epoch": 0.25634438465544174,
|
| 19390 |
+
"grad_norm": 0.15625,
|
| 19391 |
+
"learning_rate": 0.02,
|
| 19392 |
+
"loss": 1.6997,
|
| 19393 |
+
"step": 2769
|
| 19394 |
+
},
|
| 19395 |
+
{
|
| 19396 |
+
"epoch": 0.2564369611757218,
|
| 19397 |
+
"grad_norm": 0.142578125,
|
| 19398 |
+
"learning_rate": 0.02,
|
| 19399 |
+
"loss": 1.6429,
|
| 19400 |
+
"step": 2770
|
| 19401 |
+
},
|
| 19402 |
+
{
|
| 19403 |
+
"epoch": 0.25652953769600184,
|
| 19404 |
+
"grad_norm": 0.1572265625,
|
| 19405 |
+
"learning_rate": 0.02,
|
| 19406 |
+
"loss": 1.6529,
|
| 19407 |
+
"step": 2771
|
| 19408 |
+
},
|
| 19409 |
+
{
|
| 19410 |
+
"epoch": 0.2566221142162819,
|
| 19411 |
+
"grad_norm": 0.1572265625,
|
| 19412 |
+
"learning_rate": 0.02,
|
| 19413 |
+
"loss": 1.6278,
|
| 19414 |
+
"step": 2772
|
| 19415 |
+
},
|
| 19416 |
+
{
|
| 19417 |
+
"epoch": 0.25671469073656195,
|
| 19418 |
+
"grad_norm": 0.16015625,
|
| 19419 |
+
"learning_rate": 0.02,
|
| 19420 |
+
"loss": 1.6645,
|
| 19421 |
+
"step": 2773
|
| 19422 |
+
},
|
| 19423 |
+
{
|
| 19424 |
+
"epoch": 0.256807267256842,
|
| 19425 |
+
"grad_norm": 0.1513671875,
|
| 19426 |
+
"learning_rate": 0.02,
|
| 19427 |
+
"loss": 1.6599,
|
| 19428 |
+
"step": 2774
|
| 19429 |
+
},
|
| 19430 |
+
{
|
| 19431 |
+
"epoch": 0.25689984377712205,
|
| 19432 |
+
"grad_norm": 0.154296875,
|
| 19433 |
+
"learning_rate": 0.02,
|
| 19434 |
+
"loss": 1.6418,
|
| 19435 |
+
"step": 2775
|
| 19436 |
+
},
|
| 19437 |
+
{
|
| 19438 |
+
"epoch": 0.25699242029740205,
|
| 19439 |
+
"grad_norm": 0.1484375,
|
| 19440 |
+
"learning_rate": 0.02,
|
| 19441 |
+
"loss": 1.6788,
|
| 19442 |
+
"step": 2776
|
| 19443 |
+
},
|
| 19444 |
+
{
|
| 19445 |
+
"epoch": 0.2570849968176821,
|
| 19446 |
+
"grad_norm": 0.154296875,
|
| 19447 |
+
"learning_rate": 0.02,
|
| 19448 |
+
"loss": 1.638,
|
| 19449 |
+
"step": 2777
|
| 19450 |
+
},
|
| 19451 |
+
{
|
| 19452 |
+
"epoch": 0.25717757333796215,
|
| 19453 |
+
"grad_norm": 0.1484375,
|
| 19454 |
+
"learning_rate": 0.02,
|
| 19455 |
+
"loss": 1.6118,
|
| 19456 |
+
"step": 2778
|
| 19457 |
+
},
|
| 19458 |
+
{
|
| 19459 |
+
"epoch": 0.2572701498582422,
|
| 19460 |
+
"grad_norm": 0.1494140625,
|
| 19461 |
+
"learning_rate": 0.02,
|
| 19462 |
+
"loss": 1.6631,
|
| 19463 |
+
"step": 2779
|
| 19464 |
+
},
|
| 19465 |
+
{
|
| 19466 |
+
"epoch": 0.25736272637852226,
|
| 19467 |
+
"grad_norm": 0.15234375,
|
| 19468 |
+
"learning_rate": 0.02,
|
| 19469 |
+
"loss": 1.6593,
|
| 19470 |
+
"step": 2780
|
| 19471 |
+
},
|
| 19472 |
+
{
|
| 19473 |
+
"epoch": 0.2574553028988023,
|
| 19474 |
+
"grad_norm": 0.1572265625,
|
| 19475 |
+
"learning_rate": 0.02,
|
| 19476 |
+
"loss": 1.7023,
|
| 19477 |
+
"step": 2781
|
| 19478 |
+
},
|
| 19479 |
+
{
|
| 19480 |
+
"epoch": 0.25754787941908236,
|
| 19481 |
+
"grad_norm": 0.1484375,
|
| 19482 |
+
"learning_rate": 0.02,
|
| 19483 |
+
"loss": 1.6811,
|
| 19484 |
+
"step": 2782
|
| 19485 |
+
},
|
| 19486 |
+
{
|
| 19487 |
+
"epoch": 0.25764045593936236,
|
| 19488 |
+
"grad_norm": 0.142578125,
|
| 19489 |
+
"learning_rate": 0.02,
|
| 19490 |
+
"loss": 1.6432,
|
| 19491 |
+
"step": 2783
|
| 19492 |
+
},
|
| 19493 |
+
{
|
| 19494 |
+
"epoch": 0.2577330324596424,
|
| 19495 |
+
"grad_norm": 0.150390625,
|
| 19496 |
+
"learning_rate": 0.02,
|
| 19497 |
+
"loss": 1.663,
|
| 19498 |
+
"step": 2784
|
| 19499 |
+
},
|
| 19500 |
+
{
|
| 19501 |
+
"epoch": 0.25782560897992246,
|
| 19502 |
+
"grad_norm": 0.142578125,
|
| 19503 |
+
"learning_rate": 0.02,
|
| 19504 |
+
"loss": 1.5793,
|
| 19505 |
+
"step": 2785
|
| 19506 |
+
},
|
| 19507 |
+
{
|
| 19508 |
+
"epoch": 0.2579181855002025,
|
| 19509 |
+
"grad_norm": 0.1494140625,
|
| 19510 |
+
"learning_rate": 0.02,
|
| 19511 |
+
"loss": 1.6421,
|
| 19512 |
+
"step": 2786
|
| 19513 |
+
},
|
| 19514 |
+
{
|
| 19515 |
+
"epoch": 0.25801076202048256,
|
| 19516 |
+
"grad_norm": 0.150390625,
|
| 19517 |
+
"learning_rate": 0.02,
|
| 19518 |
+
"loss": 1.6715,
|
| 19519 |
+
"step": 2787
|
| 19520 |
+
},
|
| 19521 |
+
{
|
| 19522 |
+
"epoch": 0.2581033385407626,
|
| 19523 |
+
"grad_norm": 0.1396484375,
|
| 19524 |
+
"learning_rate": 0.02,
|
| 19525 |
+
"loss": 1.7028,
|
| 19526 |
+
"step": 2788
|
| 19527 |
+
},
|
| 19528 |
+
{
|
| 19529 |
+
"epoch": 0.25819591506104267,
|
| 19530 |
+
"grad_norm": 0.150390625,
|
| 19531 |
+
"learning_rate": 0.02,
|
| 19532 |
+
"loss": 1.5634,
|
| 19533 |
+
"step": 2789
|
| 19534 |
+
},
|
| 19535 |
+
{
|
| 19536 |
+
"epoch": 0.25828849158132267,
|
| 19537 |
+
"grad_norm": 0.154296875,
|
| 19538 |
+
"learning_rate": 0.02,
|
| 19539 |
+
"loss": 1.7299,
|
| 19540 |
+
"step": 2790
|
| 19541 |
+
},
|
| 19542 |
+
{
|
| 19543 |
+
"epoch": 0.2583810681016027,
|
| 19544 |
+
"grad_norm": 0.1455078125,
|
| 19545 |
+
"learning_rate": 0.02,
|
| 19546 |
+
"loss": 1.6353,
|
| 19547 |
+
"step": 2791
|
| 19548 |
+
},
|
| 19549 |
+
{
|
| 19550 |
+
"epoch": 0.25847364462188277,
|
| 19551 |
+
"grad_norm": 0.1396484375,
|
| 19552 |
+
"learning_rate": 0.02,
|
| 19553 |
+
"loss": 1.6583,
|
| 19554 |
+
"step": 2792
|
| 19555 |
+
},
|
| 19556 |
+
{
|
| 19557 |
+
"epoch": 0.2585662211421628,
|
| 19558 |
+
"grad_norm": 0.13671875,
|
| 19559 |
+
"learning_rate": 0.02,
|
| 19560 |
+
"loss": 1.6632,
|
| 19561 |
+
"step": 2793
|
| 19562 |
+
},
|
| 19563 |
+
{
|
| 19564 |
+
"epoch": 0.2586587976624429,
|
| 19565 |
+
"grad_norm": 0.1474609375,
|
| 19566 |
+
"learning_rate": 0.02,
|
| 19567 |
+
"loss": 1.7041,
|
| 19568 |
+
"step": 2794
|
| 19569 |
+
},
|
| 19570 |
+
{
|
| 19571 |
+
"epoch": 0.2587513741827229,
|
| 19572 |
+
"grad_norm": 0.1474609375,
|
| 19573 |
+
"learning_rate": 0.02,
|
| 19574 |
+
"loss": 1.6851,
|
| 19575 |
+
"step": 2795
|
| 19576 |
+
},
|
| 19577 |
+
{
|
| 19578 |
+
"epoch": 0.258843950703003,
|
| 19579 |
+
"grad_norm": 0.1416015625,
|
| 19580 |
+
"learning_rate": 0.02,
|
| 19581 |
+
"loss": 1.6089,
|
| 19582 |
+
"step": 2796
|
| 19583 |
+
},
|
| 19584 |
+
{
|
| 19585 |
+
"epoch": 0.258936527223283,
|
| 19586 |
+
"grad_norm": 0.154296875,
|
| 19587 |
+
"learning_rate": 0.02,
|
| 19588 |
+
"loss": 1.6764,
|
| 19589 |
+
"step": 2797
|
| 19590 |
+
},
|
| 19591 |
+
{
|
| 19592 |
+
"epoch": 0.259029103743563,
|
| 19593 |
+
"grad_norm": 0.1455078125,
|
| 19594 |
+
"learning_rate": 0.02,
|
| 19595 |
+
"loss": 1.5725,
|
| 19596 |
+
"step": 2798
|
| 19597 |
+
},
|
| 19598 |
+
{
|
| 19599 |
+
"epoch": 0.2591216802638431,
|
| 19600 |
+
"grad_norm": 0.1513671875,
|
| 19601 |
+
"learning_rate": 0.02,
|
| 19602 |
+
"loss": 1.6125,
|
| 19603 |
+
"step": 2799
|
| 19604 |
+
},
|
| 19605 |
+
{
|
| 19606 |
+
"epoch": 0.25921425678412313,
|
| 19607 |
+
"grad_norm": 0.154296875,
|
| 19608 |
+
"learning_rate": 0.02,
|
| 19609 |
+
"loss": 1.6711,
|
| 19610 |
+
"step": 2800
|
| 19611 |
+
},
|
| 19612 |
+
{
|
| 19613 |
+
"epoch": 0.2593068333044032,
|
| 19614 |
+
"grad_norm": 0.154296875,
|
| 19615 |
+
"learning_rate": 0.02,
|
| 19616 |
+
"loss": 1.6687,
|
| 19617 |
+
"step": 2801
|
| 19618 |
+
},
|
| 19619 |
+
{
|
| 19620 |
+
"epoch": 0.25939940982468324,
|
| 19621 |
+
"grad_norm": 0.154296875,
|
| 19622 |
+
"learning_rate": 0.02,
|
| 19623 |
+
"loss": 1.6241,
|
| 19624 |
+
"step": 2802
|
| 19625 |
+
},
|
| 19626 |
+
{
|
| 19627 |
+
"epoch": 0.25949198634496323,
|
| 19628 |
+
"grad_norm": 0.1484375,
|
| 19629 |
+
"learning_rate": 0.02,
|
| 19630 |
+
"loss": 1.6478,
|
| 19631 |
+
"step": 2803
|
| 19632 |
+
},
|
| 19633 |
+
{
|
| 19634 |
+
"epoch": 0.2595845628652433,
|
| 19635 |
+
"grad_norm": 0.1455078125,
|
| 19636 |
+
"learning_rate": 0.02,
|
| 19637 |
+
"loss": 1.6141,
|
| 19638 |
+
"step": 2804
|
| 19639 |
+
},
|
| 19640 |
+
{
|
| 19641 |
+
"epoch": 0.25967713938552334,
|
| 19642 |
+
"grad_norm": 0.1513671875,
|
| 19643 |
+
"learning_rate": 0.02,
|
| 19644 |
+
"loss": 1.6976,
|
| 19645 |
+
"step": 2805
|
| 19646 |
+
},
|
| 19647 |
+
{
|
| 19648 |
+
"epoch": 0.2597697159058034,
|
| 19649 |
+
"grad_norm": 0.1484375,
|
| 19650 |
+
"learning_rate": 0.02,
|
| 19651 |
+
"loss": 1.6385,
|
| 19652 |
+
"step": 2806
|
| 19653 |
+
},
|
| 19654 |
+
{
|
| 19655 |
+
"epoch": 0.25986229242608344,
|
| 19656 |
+
"grad_norm": 0.15234375,
|
| 19657 |
+
"learning_rate": 0.02,
|
| 19658 |
+
"loss": 1.6946,
|
| 19659 |
+
"step": 2807
|
| 19660 |
+
},
|
| 19661 |
+
{
|
| 19662 |
+
"epoch": 0.2599548689463635,
|
| 19663 |
+
"grad_norm": 0.1484375,
|
| 19664 |
+
"learning_rate": 0.02,
|
| 19665 |
+
"loss": 1.6303,
|
| 19666 |
+
"step": 2808
|
| 19667 |
+
},
|
| 19668 |
+
{
|
| 19669 |
+
"epoch": 0.26004744546664355,
|
| 19670 |
+
"grad_norm": 0.1591796875,
|
| 19671 |
+
"learning_rate": 0.02,
|
| 19672 |
+
"loss": 1.7219,
|
| 19673 |
+
"step": 2809
|
| 19674 |
+
},
|
| 19675 |
+
{
|
| 19676 |
+
"epoch": 0.26014002198692354,
|
| 19677 |
+
"grad_norm": 0.158203125,
|
| 19678 |
+
"learning_rate": 0.02,
|
| 19679 |
+
"loss": 1.6885,
|
| 19680 |
+
"step": 2810
|
| 19681 |
+
},
|
| 19682 |
+
{
|
| 19683 |
+
"epoch": 0.2602325985072036,
|
| 19684 |
+
"grad_norm": 0.1513671875,
|
| 19685 |
+
"learning_rate": 0.02,
|
| 19686 |
+
"loss": 1.6945,
|
| 19687 |
+
"step": 2811
|
| 19688 |
+
},
|
| 19689 |
+
{
|
| 19690 |
+
"epoch": 0.26032517502748365,
|
| 19691 |
+
"grad_norm": 0.1435546875,
|
| 19692 |
+
"learning_rate": 0.02,
|
| 19693 |
+
"loss": 1.6883,
|
| 19694 |
+
"step": 2812
|
| 19695 |
+
},
|
| 19696 |
+
{
|
| 19697 |
+
"epoch": 0.2604177515477637,
|
| 19698 |
+
"grad_norm": 0.15625,
|
| 19699 |
+
"learning_rate": 0.02,
|
| 19700 |
+
"loss": 1.7024,
|
| 19701 |
+
"step": 2813
|
| 19702 |
+
},
|
| 19703 |
+
{
|
| 19704 |
+
"epoch": 0.26051032806804375,
|
| 19705 |
+
"grad_norm": 0.154296875,
|
| 19706 |
+
"learning_rate": 0.02,
|
| 19707 |
+
"loss": 1.7098,
|
| 19708 |
+
"step": 2814
|
| 19709 |
+
},
|
| 19710 |
+
{
|
| 19711 |
+
"epoch": 0.2606029045883238,
|
| 19712 |
+
"grad_norm": 0.1455078125,
|
| 19713 |
+
"learning_rate": 0.02,
|
| 19714 |
+
"loss": 1.6763,
|
| 19715 |
+
"step": 2815
|
| 19716 |
+
},
|
| 19717 |
+
{
|
| 19718 |
+
"epoch": 0.26069548110860385,
|
| 19719 |
+
"grad_norm": 0.146484375,
|
| 19720 |
+
"learning_rate": 0.02,
|
| 19721 |
+
"loss": 1.6422,
|
| 19722 |
+
"step": 2816
|
| 19723 |
+
},
|
| 19724 |
+
{
|
| 19725 |
+
"epoch": 0.26078805762888385,
|
| 19726 |
+
"grad_norm": 0.140625,
|
| 19727 |
+
"learning_rate": 0.02,
|
| 19728 |
+
"loss": 1.6549,
|
| 19729 |
+
"step": 2817
|
| 19730 |
+
},
|
| 19731 |
+
{
|
| 19732 |
+
"epoch": 0.2608806341491639,
|
| 19733 |
+
"grad_norm": 0.1474609375,
|
| 19734 |
+
"learning_rate": 0.02,
|
| 19735 |
+
"loss": 1.6597,
|
| 19736 |
+
"step": 2818
|
| 19737 |
+
},
|
| 19738 |
+
{
|
| 19739 |
+
"epoch": 0.26097321066944396,
|
| 19740 |
+
"grad_norm": 0.1435546875,
|
| 19741 |
+
"learning_rate": 0.02,
|
| 19742 |
+
"loss": 1.632,
|
| 19743 |
+
"step": 2819
|
| 19744 |
+
},
|
| 19745 |
+
{
|
| 19746 |
+
"epoch": 0.261065787189724,
|
| 19747 |
+
"grad_norm": 0.1494140625,
|
| 19748 |
+
"learning_rate": 0.02,
|
| 19749 |
+
"loss": 1.7053,
|
| 19750 |
+
"step": 2820
|
| 19751 |
+
},
|
| 19752 |
+
{
|
| 19753 |
+
"epoch": 0.26115836371000406,
|
| 19754 |
+
"grad_norm": 0.140625,
|
| 19755 |
+
"learning_rate": 0.02,
|
| 19756 |
+
"loss": 1.6298,
|
| 19757 |
+
"step": 2821
|
| 19758 |
+
},
|
| 19759 |
+
{
|
| 19760 |
+
"epoch": 0.2612509402302841,
|
| 19761 |
+
"grad_norm": 0.1435546875,
|
| 19762 |
+
"learning_rate": 0.02,
|
| 19763 |
+
"loss": 1.6617,
|
| 19764 |
+
"step": 2822
|
| 19765 |
+
},
|
| 19766 |
+
{
|
| 19767 |
+
"epoch": 0.26134351675056416,
|
| 19768 |
+
"grad_norm": 0.1455078125,
|
| 19769 |
+
"learning_rate": 0.02,
|
| 19770 |
+
"loss": 1.6142,
|
| 19771 |
+
"step": 2823
|
| 19772 |
+
},
|
| 19773 |
+
{
|
| 19774 |
+
"epoch": 0.26143609327084416,
|
| 19775 |
+
"grad_norm": 0.146484375,
|
| 19776 |
+
"learning_rate": 0.02,
|
| 19777 |
+
"loss": 1.6591,
|
| 19778 |
+
"step": 2824
|
| 19779 |
+
},
|
| 19780 |
+
{
|
| 19781 |
+
"epoch": 0.2615286697911242,
|
| 19782 |
+
"grad_norm": 0.1396484375,
|
| 19783 |
+
"learning_rate": 0.02,
|
| 19784 |
+
"loss": 1.6943,
|
| 19785 |
+
"step": 2825
|
| 19786 |
+
},
|
| 19787 |
+
{
|
| 19788 |
+
"epoch": 0.26162124631140427,
|
| 19789 |
+
"grad_norm": 0.14453125,
|
| 19790 |
+
"learning_rate": 0.02,
|
| 19791 |
+
"loss": 1.6459,
|
| 19792 |
+
"step": 2826
|
| 19793 |
+
},
|
| 19794 |
+
{
|
| 19795 |
+
"epoch": 0.2617138228316843,
|
| 19796 |
+
"grad_norm": 0.1494140625,
|
| 19797 |
+
"learning_rate": 0.02,
|
| 19798 |
+
"loss": 1.7123,
|
| 19799 |
+
"step": 2827
|
| 19800 |
+
},
|
| 19801 |
+
{
|
| 19802 |
+
"epoch": 0.26180639935196437,
|
| 19803 |
+
"grad_norm": 0.1552734375,
|
| 19804 |
+
"learning_rate": 0.02,
|
| 19805 |
+
"loss": 1.7169,
|
| 19806 |
+
"step": 2828
|
| 19807 |
+
},
|
| 19808 |
+
{
|
| 19809 |
+
"epoch": 0.2618989758722444,
|
| 19810 |
+
"grad_norm": 0.1435546875,
|
| 19811 |
+
"learning_rate": 0.02,
|
| 19812 |
+
"loss": 1.6423,
|
| 19813 |
+
"step": 2829
|
| 19814 |
+
},
|
| 19815 |
+
{
|
| 19816 |
+
"epoch": 0.2619915523925244,
|
| 19817 |
+
"grad_norm": 0.134765625,
|
| 19818 |
+
"learning_rate": 0.02,
|
| 19819 |
+
"loss": 1.6434,
|
| 19820 |
+
"step": 2830
|
| 19821 |
+
},
|
| 19822 |
+
{
|
| 19823 |
+
"epoch": 0.26208412891280447,
|
| 19824 |
+
"grad_norm": 0.1494140625,
|
| 19825 |
+
"learning_rate": 0.02,
|
| 19826 |
+
"loss": 1.6682,
|
| 19827 |
+
"step": 2831
|
| 19828 |
+
},
|
| 19829 |
+
{
|
| 19830 |
+
"epoch": 0.2621767054330845,
|
| 19831 |
+
"grad_norm": 0.1552734375,
|
| 19832 |
+
"learning_rate": 0.02,
|
| 19833 |
+
"loss": 1.6019,
|
| 19834 |
+
"step": 2832
|
| 19835 |
+
},
|
| 19836 |
+
{
|
| 19837 |
+
"epoch": 0.2622692819533646,
|
| 19838 |
+
"grad_norm": 0.1552734375,
|
| 19839 |
+
"learning_rate": 0.02,
|
| 19840 |
+
"loss": 1.6233,
|
| 19841 |
+
"step": 2833
|
| 19842 |
+
},
|
| 19843 |
+
{
|
| 19844 |
+
"epoch": 0.2623618584736446,
|
| 19845 |
+
"grad_norm": 0.1552734375,
|
| 19846 |
+
"learning_rate": 0.02,
|
| 19847 |
+
"loss": 1.6967,
|
| 19848 |
+
"step": 2834
|
| 19849 |
+
},
|
| 19850 |
+
{
|
| 19851 |
+
"epoch": 0.2624544349939247,
|
| 19852 |
+
"grad_norm": 0.158203125,
|
| 19853 |
+
"learning_rate": 0.02,
|
| 19854 |
+
"loss": 1.6257,
|
| 19855 |
+
"step": 2835
|
| 19856 |
+
},
|
| 19857 |
+
{
|
| 19858 |
+
"epoch": 0.26254701151420473,
|
| 19859 |
+
"grad_norm": 0.15234375,
|
| 19860 |
+
"learning_rate": 0.02,
|
| 19861 |
+
"loss": 1.6696,
|
| 19862 |
+
"step": 2836
|
| 19863 |
+
},
|
| 19864 |
+
{
|
| 19865 |
+
"epoch": 0.26263958803448473,
|
| 19866 |
+
"grad_norm": 0.158203125,
|
| 19867 |
+
"learning_rate": 0.02,
|
| 19868 |
+
"loss": 1.6953,
|
| 19869 |
+
"step": 2837
|
| 19870 |
+
},
|
| 19871 |
+
{
|
| 19872 |
+
"epoch": 0.2627321645547648,
|
| 19873 |
+
"grad_norm": 0.1474609375,
|
| 19874 |
+
"learning_rate": 0.02,
|
| 19875 |
+
"loss": 1.6865,
|
| 19876 |
+
"step": 2838
|
| 19877 |
+
},
|
| 19878 |
+
{
|
| 19879 |
+
"epoch": 0.26282474107504483,
|
| 19880 |
+
"grad_norm": 0.1513671875,
|
| 19881 |
+
"learning_rate": 0.02,
|
| 19882 |
+
"loss": 1.6295,
|
| 19883 |
+
"step": 2839
|
| 19884 |
+
},
|
| 19885 |
+
{
|
| 19886 |
+
"epoch": 0.2629173175953249,
|
| 19887 |
+
"grad_norm": 0.1591796875,
|
| 19888 |
+
"learning_rate": 0.02,
|
| 19889 |
+
"loss": 1.6418,
|
| 19890 |
+
"step": 2840
|
| 19891 |
+
},
|
| 19892 |
+
{
|
| 19893 |
+
"epoch": 0.26300989411560494,
|
| 19894 |
+
"grad_norm": 0.1572265625,
|
| 19895 |
+
"learning_rate": 0.02,
|
| 19896 |
+
"loss": 1.6713,
|
| 19897 |
+
"step": 2841
|
| 19898 |
+
},
|
| 19899 |
+
{
|
| 19900 |
+
"epoch": 0.263102470635885,
|
| 19901 |
+
"grad_norm": 0.1474609375,
|
| 19902 |
+
"learning_rate": 0.02,
|
| 19903 |
+
"loss": 1.6951,
|
| 19904 |
+
"step": 2842
|
| 19905 |
+
},
|
| 19906 |
+
{
|
| 19907 |
+
"epoch": 0.26319504715616504,
|
| 19908 |
+
"grad_norm": 0.1474609375,
|
| 19909 |
+
"learning_rate": 0.02,
|
| 19910 |
+
"loss": 1.6727,
|
| 19911 |
+
"step": 2843
|
| 19912 |
+
},
|
| 19913 |
+
{
|
| 19914 |
+
"epoch": 0.26328762367644504,
|
| 19915 |
+
"grad_norm": 0.146484375,
|
| 19916 |
+
"learning_rate": 0.02,
|
| 19917 |
+
"loss": 1.5912,
|
| 19918 |
+
"step": 2844
|
| 19919 |
+
},
|
| 19920 |
+
{
|
| 19921 |
+
"epoch": 0.2633802001967251,
|
| 19922 |
+
"grad_norm": 0.1533203125,
|
| 19923 |
+
"learning_rate": 0.02,
|
| 19924 |
+
"loss": 1.6671,
|
| 19925 |
+
"step": 2845
|
| 19926 |
+
},
|
| 19927 |
+
{
|
| 19928 |
+
"epoch": 0.26347277671700514,
|
| 19929 |
+
"grad_norm": 0.1533203125,
|
| 19930 |
+
"learning_rate": 0.02,
|
| 19931 |
+
"loss": 1.6898,
|
| 19932 |
+
"step": 2846
|
| 19933 |
+
},
|
| 19934 |
+
{
|
| 19935 |
+
"epoch": 0.2635653532372852,
|
| 19936 |
+
"grad_norm": 0.1591796875,
|
| 19937 |
+
"learning_rate": 0.02,
|
| 19938 |
+
"loss": 1.6778,
|
| 19939 |
+
"step": 2847
|
| 19940 |
+
},
|
| 19941 |
+
{
|
| 19942 |
+
"epoch": 0.26365792975756525,
|
| 19943 |
+
"grad_norm": 0.1484375,
|
| 19944 |
+
"learning_rate": 0.02,
|
| 19945 |
+
"loss": 1.6392,
|
| 19946 |
+
"step": 2848
|
| 19947 |
+
},
|
| 19948 |
+
{
|
| 19949 |
+
"epoch": 0.2637505062778453,
|
| 19950 |
+
"grad_norm": 0.1591796875,
|
| 19951 |
+
"learning_rate": 0.02,
|
| 19952 |
+
"loss": 1.6483,
|
| 19953 |
+
"step": 2849
|
| 19954 |
+
},
|
| 19955 |
+
{
|
| 19956 |
+
"epoch": 0.26384308279812535,
|
| 19957 |
+
"grad_norm": 0.154296875,
|
| 19958 |
+
"learning_rate": 0.02,
|
| 19959 |
+
"loss": 1.6028,
|
| 19960 |
+
"step": 2850
|
| 19961 |
+
},
|
| 19962 |
+
{
|
| 19963 |
+
"epoch": 0.26393565931840535,
|
| 19964 |
+
"grad_norm": 0.16015625,
|
| 19965 |
+
"learning_rate": 0.02,
|
| 19966 |
+
"loss": 1.5759,
|
| 19967 |
+
"step": 2851
|
| 19968 |
+
},
|
| 19969 |
+
{
|
| 19970 |
+
"epoch": 0.2640282358386854,
|
| 19971 |
+
"grad_norm": 0.16015625,
|
| 19972 |
+
"learning_rate": 0.02,
|
| 19973 |
+
"loss": 1.6332,
|
| 19974 |
+
"step": 2852
|
| 19975 |
+
},
|
| 19976 |
+
{
|
| 19977 |
+
"epoch": 0.26412081235896545,
|
| 19978 |
+
"grad_norm": 0.15625,
|
| 19979 |
+
"learning_rate": 0.02,
|
| 19980 |
+
"loss": 1.6672,
|
| 19981 |
+
"step": 2853
|
| 19982 |
+
},
|
| 19983 |
+
{
|
| 19984 |
+
"epoch": 0.2642133888792455,
|
| 19985 |
+
"grad_norm": 0.1611328125,
|
| 19986 |
+
"learning_rate": 0.02,
|
| 19987 |
+
"loss": 1.6469,
|
| 19988 |
+
"step": 2854
|
| 19989 |
+
},
|
| 19990 |
+
{
|
| 19991 |
+
"epoch": 0.26430596539952556,
|
| 19992 |
+
"grad_norm": 0.1513671875,
|
| 19993 |
+
"learning_rate": 0.02,
|
| 19994 |
+
"loss": 1.7041,
|
| 19995 |
+
"step": 2855
|
| 19996 |
+
},
|
| 19997 |
+
{
|
| 19998 |
+
"epoch": 0.2643985419198056,
|
| 19999 |
+
"grad_norm": 0.1494140625,
|
| 20000 |
+
"learning_rate": 0.02,
|
| 20001 |
+
"loss": 1.6714,
|
| 20002 |
+
"step": 2856
|
| 20003 |
+
},
|
| 20004 |
+
{
|
| 20005 |
+
"epoch": 0.26449111844008566,
|
| 20006 |
+
"grad_norm": 0.142578125,
|
| 20007 |
+
"learning_rate": 0.02,
|
| 20008 |
+
"loss": 1.5934,
|
| 20009 |
+
"step": 2857
|
| 20010 |
+
},
|
| 20011 |
+
{
|
| 20012 |
+
"epoch": 0.26458369496036566,
|
| 20013 |
+
"grad_norm": 0.15234375,
|
| 20014 |
+
"learning_rate": 0.02,
|
| 20015 |
+
"loss": 1.6739,
|
| 20016 |
+
"step": 2858
|
| 20017 |
+
},
|
| 20018 |
+
{
|
| 20019 |
+
"epoch": 0.2646762714806457,
|
| 20020 |
+
"grad_norm": 0.1474609375,
|
| 20021 |
+
"learning_rate": 0.02,
|
| 20022 |
+
"loss": 1.6576,
|
| 20023 |
+
"step": 2859
|
| 20024 |
+
},
|
| 20025 |
+
{
|
| 20026 |
+
"epoch": 0.26476884800092576,
|
| 20027 |
+
"grad_norm": 0.154296875,
|
| 20028 |
+
"learning_rate": 0.02,
|
| 20029 |
+
"loss": 1.6805,
|
| 20030 |
+
"step": 2860
|
| 20031 |
+
},
|
| 20032 |
+
{
|
| 20033 |
+
"epoch": 0.2648614245212058,
|
| 20034 |
+
"grad_norm": 0.1416015625,
|
| 20035 |
+
"learning_rate": 0.02,
|
| 20036 |
+
"loss": 1.6645,
|
| 20037 |
+
"step": 2861
|
| 20038 |
+
},
|
| 20039 |
+
{
|
| 20040 |
+
"epoch": 0.26495400104148586,
|
| 20041 |
+
"grad_norm": 0.15234375,
|
| 20042 |
+
"learning_rate": 0.02,
|
| 20043 |
+
"loss": 1.6594,
|
| 20044 |
+
"step": 2862
|
| 20045 |
+
},
|
| 20046 |
+
{
|
| 20047 |
+
"epoch": 0.2650465775617659,
|
| 20048 |
+
"grad_norm": 0.146484375,
|
| 20049 |
+
"learning_rate": 0.02,
|
| 20050 |
+
"loss": 1.6264,
|
| 20051 |
+
"step": 2863
|
| 20052 |
+
},
|
| 20053 |
+
{
|
| 20054 |
+
"epoch": 0.2651391540820459,
|
| 20055 |
+
"grad_norm": 0.1455078125,
|
| 20056 |
+
"learning_rate": 0.02,
|
| 20057 |
+
"loss": 1.6439,
|
| 20058 |
+
"step": 2864
|
| 20059 |
+
},
|
| 20060 |
+
{
|
| 20061 |
+
"epoch": 0.26523173060232597,
|
| 20062 |
+
"grad_norm": 0.1396484375,
|
| 20063 |
+
"learning_rate": 0.02,
|
| 20064 |
+
"loss": 1.6258,
|
| 20065 |
+
"step": 2865
|
| 20066 |
+
},
|
| 20067 |
+
{
|
| 20068 |
+
"epoch": 0.265324307122606,
|
| 20069 |
+
"grad_norm": 0.150390625,
|
| 20070 |
+
"learning_rate": 0.02,
|
| 20071 |
+
"loss": 1.6453,
|
| 20072 |
+
"step": 2866
|
| 20073 |
+
},
|
| 20074 |
+
{
|
| 20075 |
+
"epoch": 0.26541688364288607,
|
| 20076 |
+
"grad_norm": 0.150390625,
|
| 20077 |
+
"learning_rate": 0.02,
|
| 20078 |
+
"loss": 1.643,
|
| 20079 |
+
"step": 2867
|
| 20080 |
+
},
|
| 20081 |
+
{
|
| 20082 |
+
"epoch": 0.2655094601631661,
|
| 20083 |
+
"grad_norm": 0.1474609375,
|
| 20084 |
+
"learning_rate": 0.02,
|
| 20085 |
+
"loss": 1.7105,
|
| 20086 |
+
"step": 2868
|
| 20087 |
+
},
|
| 20088 |
+
{
|
| 20089 |
+
"epoch": 0.2656020366834462,
|
| 20090 |
+
"grad_norm": 0.1591796875,
|
| 20091 |
+
"learning_rate": 0.02,
|
| 20092 |
+
"loss": 1.6835,
|
| 20093 |
+
"step": 2869
|
| 20094 |
+
},
|
| 20095 |
+
{
|
| 20096 |
+
"epoch": 0.2656946132037262,
|
| 20097 |
+
"grad_norm": 0.1484375,
|
| 20098 |
+
"learning_rate": 0.02,
|
| 20099 |
+
"loss": 1.6739,
|
| 20100 |
+
"step": 2870
|
| 20101 |
+
},
|
| 20102 |
+
{
|
| 20103 |
+
"epoch": 0.2657871897240062,
|
| 20104 |
+
"grad_norm": 0.154296875,
|
| 20105 |
+
"learning_rate": 0.02,
|
| 20106 |
+
"loss": 1.6796,
|
| 20107 |
+
"step": 2871
|
| 20108 |
+
},
|
| 20109 |
+
{
|
| 20110 |
+
"epoch": 0.2658797662442863,
|
| 20111 |
+
"grad_norm": 0.146484375,
|
| 20112 |
+
"learning_rate": 0.02,
|
| 20113 |
+
"loss": 1.6223,
|
| 20114 |
+
"step": 2872
|
| 20115 |
+
},
|
| 20116 |
+
{
|
| 20117 |
+
"epoch": 0.2659723427645663,
|
| 20118 |
+
"grad_norm": 0.1455078125,
|
| 20119 |
+
"learning_rate": 0.02,
|
| 20120 |
+
"loss": 1.6662,
|
| 20121 |
+
"step": 2873
|
| 20122 |
+
},
|
| 20123 |
+
{
|
| 20124 |
+
"epoch": 0.2660649192848464,
|
| 20125 |
+
"grad_norm": 0.1455078125,
|
| 20126 |
+
"learning_rate": 0.02,
|
| 20127 |
+
"loss": 1.6816,
|
| 20128 |
+
"step": 2874
|
| 20129 |
+
},
|
| 20130 |
+
{
|
| 20131 |
+
"epoch": 0.26615749580512643,
|
| 20132 |
+
"grad_norm": 0.1513671875,
|
| 20133 |
+
"learning_rate": 0.02,
|
| 20134 |
+
"loss": 1.6126,
|
| 20135 |
+
"step": 2875
|
| 20136 |
+
},
|
| 20137 |
+
{
|
| 20138 |
+
"epoch": 0.2662500723254065,
|
| 20139 |
+
"grad_norm": 0.1572265625,
|
| 20140 |
+
"learning_rate": 0.02,
|
| 20141 |
+
"loss": 1.719,
|
| 20142 |
+
"step": 2876
|
| 20143 |
+
},
|
| 20144 |
+
{
|
| 20145 |
+
"epoch": 0.26634264884568654,
|
| 20146 |
+
"grad_norm": 0.15234375,
|
| 20147 |
+
"learning_rate": 0.02,
|
| 20148 |
+
"loss": 1.639,
|
| 20149 |
+
"step": 2877
|
| 20150 |
+
},
|
| 20151 |
+
{
|
| 20152 |
+
"epoch": 0.26643522536596653,
|
| 20153 |
+
"grad_norm": 0.154296875,
|
| 20154 |
+
"learning_rate": 0.02,
|
| 20155 |
+
"loss": 1.6367,
|
| 20156 |
+
"step": 2878
|
| 20157 |
+
},
|
| 20158 |
+
{
|
| 20159 |
+
"epoch": 0.2665278018862466,
|
| 20160 |
+
"grad_norm": 0.1494140625,
|
| 20161 |
+
"learning_rate": 0.02,
|
| 20162 |
+
"loss": 1.5886,
|
| 20163 |
+
"step": 2879
|
| 20164 |
+
},
|
| 20165 |
+
{
|
| 20166 |
+
"epoch": 0.26662037840652664,
|
| 20167 |
+
"grad_norm": 0.154296875,
|
| 20168 |
+
"learning_rate": 0.02,
|
| 20169 |
+
"loss": 1.5996,
|
| 20170 |
+
"step": 2880
|
| 20171 |
+
},
|
| 20172 |
+
{
|
| 20173 |
+
"epoch": 0.2667129549268067,
|
| 20174 |
+
"grad_norm": 0.1474609375,
|
| 20175 |
+
"learning_rate": 0.02,
|
| 20176 |
+
"loss": 1.6856,
|
| 20177 |
+
"step": 2881
|
| 20178 |
+
},
|
| 20179 |
+
{
|
| 20180 |
+
"epoch": 0.26680553144708674,
|
| 20181 |
+
"grad_norm": 0.16015625,
|
| 20182 |
+
"learning_rate": 0.02,
|
| 20183 |
+
"loss": 1.7201,
|
| 20184 |
+
"step": 2882
|
| 20185 |
+
},
|
| 20186 |
+
{
|
| 20187 |
+
"epoch": 0.2668981079673668,
|
| 20188 |
+
"grad_norm": 0.1474609375,
|
| 20189 |
+
"learning_rate": 0.02,
|
| 20190 |
+
"loss": 1.6602,
|
| 20191 |
+
"step": 2883
|
| 20192 |
+
},
|
| 20193 |
+
{
|
| 20194 |
+
"epoch": 0.26699068448764685,
|
| 20195 |
+
"grad_norm": 0.1611328125,
|
| 20196 |
+
"learning_rate": 0.02,
|
| 20197 |
+
"loss": 1.7035,
|
| 20198 |
+
"step": 2884
|
| 20199 |
+
},
|
| 20200 |
+
{
|
| 20201 |
+
"epoch": 0.26708326100792684,
|
| 20202 |
+
"grad_norm": 0.15625,
|
| 20203 |
+
"learning_rate": 0.02,
|
| 20204 |
+
"loss": 1.6811,
|
| 20205 |
+
"step": 2885
|
| 20206 |
+
},
|
| 20207 |
+
{
|
| 20208 |
+
"epoch": 0.2671758375282069,
|
| 20209 |
+
"grad_norm": 0.16015625,
|
| 20210 |
+
"learning_rate": 0.02,
|
| 20211 |
+
"loss": 1.6841,
|
| 20212 |
+
"step": 2886
|
| 20213 |
+
},
|
| 20214 |
+
{
|
| 20215 |
+
"epoch": 0.26726841404848695,
|
| 20216 |
+
"grad_norm": 0.1435546875,
|
| 20217 |
+
"learning_rate": 0.02,
|
| 20218 |
+
"loss": 1.6108,
|
| 20219 |
+
"step": 2887
|
| 20220 |
+
},
|
| 20221 |
+
{
|
| 20222 |
+
"epoch": 0.267360990568767,
|
| 20223 |
+
"grad_norm": 0.15625,
|
| 20224 |
+
"learning_rate": 0.02,
|
| 20225 |
+
"loss": 1.7174,
|
| 20226 |
+
"step": 2888
|
| 20227 |
+
},
|
| 20228 |
+
{
|
| 20229 |
+
"epoch": 0.26745356708904705,
|
| 20230 |
+
"grad_norm": 0.1435546875,
|
| 20231 |
+
"learning_rate": 0.02,
|
| 20232 |
+
"loss": 1.5924,
|
| 20233 |
+
"step": 2889
|
| 20234 |
+
},
|
| 20235 |
+
{
|
| 20236 |
+
"epoch": 0.2675461436093271,
|
| 20237 |
+
"grad_norm": 0.1513671875,
|
| 20238 |
+
"learning_rate": 0.02,
|
| 20239 |
+
"loss": 1.6595,
|
| 20240 |
+
"step": 2890
|
| 20241 |
+
},
|
| 20242 |
+
{
|
| 20243 |
+
"epoch": 0.26763872012960715,
|
| 20244 |
+
"grad_norm": 0.1513671875,
|
| 20245 |
+
"learning_rate": 0.02,
|
| 20246 |
+
"loss": 1.6713,
|
| 20247 |
+
"step": 2891
|
| 20248 |
+
},
|
| 20249 |
+
{
|
| 20250 |
+
"epoch": 0.26773129664988715,
|
| 20251 |
+
"grad_norm": 0.162109375,
|
| 20252 |
+
"learning_rate": 0.02,
|
| 20253 |
+
"loss": 1.7279,
|
| 20254 |
+
"step": 2892
|
| 20255 |
+
},
|
| 20256 |
+
{
|
| 20257 |
+
"epoch": 0.2678238731701672,
|
| 20258 |
+
"grad_norm": 0.1455078125,
|
| 20259 |
+
"learning_rate": 0.02,
|
| 20260 |
+
"loss": 1.6218,
|
| 20261 |
+
"step": 2893
|
| 20262 |
+
},
|
| 20263 |
+
{
|
| 20264 |
+
"epoch": 0.26791644969044726,
|
| 20265 |
+
"grad_norm": 0.158203125,
|
| 20266 |
+
"learning_rate": 0.02,
|
| 20267 |
+
"loss": 1.6724,
|
| 20268 |
+
"step": 2894
|
| 20269 |
+
},
|
| 20270 |
+
{
|
| 20271 |
+
"epoch": 0.2680090262107273,
|
| 20272 |
+
"grad_norm": 0.1474609375,
|
| 20273 |
+
"learning_rate": 0.02,
|
| 20274 |
+
"loss": 1.7069,
|
| 20275 |
+
"step": 2895
|
| 20276 |
+
},
|
| 20277 |
+
{
|
| 20278 |
+
"epoch": 0.26810160273100736,
|
| 20279 |
+
"grad_norm": 0.1494140625,
|
| 20280 |
+
"learning_rate": 0.02,
|
| 20281 |
+
"loss": 1.6208,
|
| 20282 |
+
"step": 2896
|
| 20283 |
+
},
|
| 20284 |
+
{
|
| 20285 |
+
"epoch": 0.2681941792512874,
|
| 20286 |
+
"grad_norm": 0.1396484375,
|
| 20287 |
+
"learning_rate": 0.02,
|
| 20288 |
+
"loss": 1.691,
|
| 20289 |
+
"step": 2897
|
| 20290 |
+
},
|
| 20291 |
+
{
|
| 20292 |
+
"epoch": 0.2682867557715674,
|
| 20293 |
+
"grad_norm": 0.1591796875,
|
| 20294 |
+
"learning_rate": 0.02,
|
| 20295 |
+
"loss": 1.7264,
|
| 20296 |
+
"step": 2898
|
| 20297 |
+
},
|
| 20298 |
+
{
|
| 20299 |
+
"epoch": 0.26837933229184746,
|
| 20300 |
+
"grad_norm": 0.1455078125,
|
| 20301 |
+
"learning_rate": 0.02,
|
| 20302 |
+
"loss": 1.6605,
|
| 20303 |
+
"step": 2899
|
| 20304 |
+
},
|
| 20305 |
+
{
|
| 20306 |
+
"epoch": 0.2684719088121275,
|
| 20307 |
+
"grad_norm": 0.150390625,
|
| 20308 |
+
"learning_rate": 0.02,
|
| 20309 |
+
"loss": 1.6441,
|
| 20310 |
+
"step": 2900
|
| 20311 |
+
},
|
| 20312 |
+
{
|
| 20313 |
+
"epoch": 0.26856448533240757,
|
| 20314 |
+
"grad_norm": 0.1494140625,
|
| 20315 |
+
"learning_rate": 0.02,
|
| 20316 |
+
"loss": 1.6617,
|
| 20317 |
+
"step": 2901
|
| 20318 |
+
},
|
| 20319 |
+
{
|
| 20320 |
+
"epoch": 0.2686570618526876,
|
| 20321 |
+
"grad_norm": 0.1552734375,
|
| 20322 |
+
"learning_rate": 0.02,
|
| 20323 |
+
"loss": 1.6914,
|
| 20324 |
+
"step": 2902
|
| 20325 |
+
},
|
| 20326 |
+
{
|
| 20327 |
+
"epoch": 0.26874963837296767,
|
| 20328 |
+
"grad_norm": 0.1513671875,
|
| 20329 |
+
"learning_rate": 0.02,
|
| 20330 |
+
"loss": 1.6822,
|
| 20331 |
+
"step": 2903
|
| 20332 |
+
},
|
| 20333 |
+
{
|
| 20334 |
+
"epoch": 0.2688422148932477,
|
| 20335 |
+
"grad_norm": 0.1376953125,
|
| 20336 |
+
"learning_rate": 0.02,
|
| 20337 |
+
"loss": 1.5771,
|
| 20338 |
+
"step": 2904
|
| 20339 |
+
},
|
| 20340 |
+
{
|
| 20341 |
+
"epoch": 0.2689347914135277,
|
| 20342 |
+
"grad_norm": 0.146484375,
|
| 20343 |
+
"learning_rate": 0.02,
|
| 20344 |
+
"loss": 1.6515,
|
| 20345 |
+
"step": 2905
|
| 20346 |
+
},
|
| 20347 |
+
{
|
| 20348 |
+
"epoch": 0.26902736793380777,
|
| 20349 |
+
"grad_norm": 0.1494140625,
|
| 20350 |
+
"learning_rate": 0.02,
|
| 20351 |
+
"loss": 1.6529,
|
| 20352 |
+
"step": 2906
|
| 20353 |
+
},
|
| 20354 |
+
{
|
| 20355 |
+
"epoch": 0.2691199444540878,
|
| 20356 |
+
"grad_norm": 0.1474609375,
|
| 20357 |
+
"learning_rate": 0.02,
|
| 20358 |
+
"loss": 1.639,
|
| 20359 |
+
"step": 2907
|
| 20360 |
+
},
|
| 20361 |
+
{
|
| 20362 |
+
"epoch": 0.2692125209743679,
|
| 20363 |
+
"grad_norm": 0.1435546875,
|
| 20364 |
+
"learning_rate": 0.02,
|
| 20365 |
+
"loss": 1.6697,
|
| 20366 |
+
"step": 2908
|
| 20367 |
+
},
|
| 20368 |
+
{
|
| 20369 |
+
"epoch": 0.2693050974946479,
|
| 20370 |
+
"grad_norm": 0.1396484375,
|
| 20371 |
+
"learning_rate": 0.02,
|
| 20372 |
+
"loss": 1.6447,
|
| 20373 |
+
"step": 2909
|
| 20374 |
+
},
|
| 20375 |
+
{
|
| 20376 |
+
"epoch": 0.269397674014928,
|
| 20377 |
+
"grad_norm": 0.142578125,
|
| 20378 |
+
"learning_rate": 0.02,
|
| 20379 |
+
"loss": 1.6552,
|
| 20380 |
+
"step": 2910
|
| 20381 |
+
},
|
| 20382 |
+
{
|
| 20383 |
+
"epoch": 0.26949025053520803,
|
| 20384 |
+
"grad_norm": 0.1494140625,
|
| 20385 |
+
"learning_rate": 0.02,
|
| 20386 |
+
"loss": 1.7034,
|
| 20387 |
+
"step": 2911
|
| 20388 |
+
},
|
| 20389 |
+
{
|
| 20390 |
+
"epoch": 0.26958282705548803,
|
| 20391 |
+
"grad_norm": 0.1484375,
|
| 20392 |
+
"learning_rate": 0.02,
|
| 20393 |
+
"loss": 1.7098,
|
| 20394 |
+
"step": 2912
|
| 20395 |
+
},
|
| 20396 |
+
{
|
| 20397 |
+
"epoch": 0.2696754035757681,
|
| 20398 |
+
"grad_norm": 0.1494140625,
|
| 20399 |
+
"learning_rate": 0.02,
|
| 20400 |
+
"loss": 1.7295,
|
| 20401 |
+
"step": 2913
|
| 20402 |
+
},
|
| 20403 |
+
{
|
| 20404 |
+
"epoch": 0.26976798009604813,
|
| 20405 |
+
"grad_norm": 0.1474609375,
|
| 20406 |
+
"learning_rate": 0.02,
|
| 20407 |
+
"loss": 1.6588,
|
| 20408 |
+
"step": 2914
|
| 20409 |
+
},
|
| 20410 |
+
{
|
| 20411 |
+
"epoch": 0.2698605566163282,
|
| 20412 |
+
"grad_norm": 0.138671875,
|
| 20413 |
+
"learning_rate": 0.02,
|
| 20414 |
+
"loss": 1.6701,
|
| 20415 |
+
"step": 2915
|
| 20416 |
+
},
|
| 20417 |
+
{
|
| 20418 |
+
"epoch": 0.26995313313660824,
|
| 20419 |
+
"grad_norm": 0.154296875,
|
| 20420 |
+
"learning_rate": 0.02,
|
| 20421 |
+
"loss": 1.6751,
|
| 20422 |
+
"step": 2916
|
| 20423 |
+
},
|
| 20424 |
+
{
|
| 20425 |
+
"epoch": 0.2700457096568883,
|
| 20426 |
+
"grad_norm": 0.1513671875,
|
| 20427 |
+
"learning_rate": 0.02,
|
| 20428 |
+
"loss": 1.6817,
|
| 20429 |
+
"step": 2917
|
| 20430 |
+
},
|
| 20431 |
+
{
|
| 20432 |
+
"epoch": 0.27013828617716834,
|
| 20433 |
+
"grad_norm": 0.1494140625,
|
| 20434 |
+
"learning_rate": 0.02,
|
| 20435 |
+
"loss": 1.7306,
|
| 20436 |
+
"step": 2918
|
| 20437 |
+
},
|
| 20438 |
+
{
|
| 20439 |
+
"epoch": 0.27023086269744834,
|
| 20440 |
+
"grad_norm": 0.142578125,
|
| 20441 |
+
"learning_rate": 0.02,
|
| 20442 |
+
"loss": 1.6562,
|
| 20443 |
+
"step": 2919
|
| 20444 |
+
},
|
| 20445 |
+
{
|
| 20446 |
+
"epoch": 0.2703234392177284,
|
| 20447 |
+
"grad_norm": 0.142578125,
|
| 20448 |
+
"learning_rate": 0.02,
|
| 20449 |
+
"loss": 1.5626,
|
| 20450 |
+
"step": 2920
|
| 20451 |
+
},
|
| 20452 |
+
{
|
| 20453 |
+
"epoch": 0.27041601573800844,
|
| 20454 |
+
"grad_norm": 0.1416015625,
|
| 20455 |
+
"learning_rate": 0.02,
|
| 20456 |
+
"loss": 1.6434,
|
| 20457 |
+
"step": 2921
|
| 20458 |
+
},
|
| 20459 |
+
{
|
| 20460 |
+
"epoch": 0.2705085922582885,
|
| 20461 |
+
"grad_norm": 0.1513671875,
|
| 20462 |
+
"learning_rate": 0.02,
|
| 20463 |
+
"loss": 1.6845,
|
| 20464 |
+
"step": 2922
|
| 20465 |
+
},
|
| 20466 |
+
{
|
| 20467 |
+
"epoch": 0.27060116877856855,
|
| 20468 |
+
"grad_norm": 0.1435546875,
|
| 20469 |
+
"learning_rate": 0.02,
|
| 20470 |
+
"loss": 1.6769,
|
| 20471 |
+
"step": 2923
|
| 20472 |
+
},
|
| 20473 |
+
{
|
| 20474 |
+
"epoch": 0.2706937452988486,
|
| 20475 |
+
"grad_norm": 0.146484375,
|
| 20476 |
+
"learning_rate": 0.02,
|
| 20477 |
+
"loss": 1.6294,
|
| 20478 |
+
"step": 2924
|
| 20479 |
+
},
|
| 20480 |
+
{
|
| 20481 |
+
"epoch": 0.27078632181912865,
|
| 20482 |
+
"grad_norm": 0.1435546875,
|
| 20483 |
+
"learning_rate": 0.02,
|
| 20484 |
+
"loss": 1.6379,
|
| 20485 |
+
"step": 2925
|
| 20486 |
+
},
|
| 20487 |
+
{
|
| 20488 |
+
"epoch": 0.27087889833940865,
|
| 20489 |
+
"grad_norm": 0.142578125,
|
| 20490 |
+
"learning_rate": 0.02,
|
| 20491 |
+
"loss": 1.6709,
|
| 20492 |
+
"step": 2926
|
| 20493 |
+
},
|
| 20494 |
+
{
|
| 20495 |
+
"epoch": 0.2709714748596887,
|
| 20496 |
+
"grad_norm": 0.1357421875,
|
| 20497 |
+
"learning_rate": 0.02,
|
| 20498 |
+
"loss": 1.586,
|
| 20499 |
+
"step": 2927
|
| 20500 |
+
},
|
| 20501 |
+
{
|
| 20502 |
+
"epoch": 0.27106405137996875,
|
| 20503 |
+
"grad_norm": 0.14453125,
|
| 20504 |
+
"learning_rate": 0.02,
|
| 20505 |
+
"loss": 1.5888,
|
| 20506 |
+
"step": 2928
|
| 20507 |
+
},
|
| 20508 |
+
{
|
| 20509 |
+
"epoch": 0.2711566279002488,
|
| 20510 |
+
"grad_norm": 0.142578125,
|
| 20511 |
+
"learning_rate": 0.02,
|
| 20512 |
+
"loss": 1.587,
|
| 20513 |
+
"step": 2929
|
| 20514 |
+
},
|
| 20515 |
+
{
|
| 20516 |
+
"epoch": 0.27124920442052886,
|
| 20517 |
+
"grad_norm": 0.146484375,
|
| 20518 |
+
"learning_rate": 0.02,
|
| 20519 |
+
"loss": 1.6487,
|
| 20520 |
+
"step": 2930
|
| 20521 |
+
},
|
| 20522 |
+
{
|
| 20523 |
+
"epoch": 0.2713417809408089,
|
| 20524 |
+
"grad_norm": 0.1474609375,
|
| 20525 |
+
"learning_rate": 0.02,
|
| 20526 |
+
"loss": 1.6153,
|
| 20527 |
+
"step": 2931
|
| 20528 |
+
},
|
| 20529 |
+
{
|
| 20530 |
+
"epoch": 0.2714343574610889,
|
| 20531 |
+
"grad_norm": 0.16015625,
|
| 20532 |
+
"learning_rate": 0.02,
|
| 20533 |
+
"loss": 1.6689,
|
| 20534 |
+
"step": 2932
|
| 20535 |
+
},
|
| 20536 |
+
{
|
| 20537 |
+
"epoch": 0.27152693398136896,
|
| 20538 |
+
"grad_norm": 0.1669921875,
|
| 20539 |
+
"learning_rate": 0.02,
|
| 20540 |
+
"loss": 1.6482,
|
| 20541 |
+
"step": 2933
|
| 20542 |
+
},
|
| 20543 |
+
{
|
| 20544 |
+
"epoch": 0.271619510501649,
|
| 20545 |
+
"grad_norm": 0.150390625,
|
| 20546 |
+
"learning_rate": 0.02,
|
| 20547 |
+
"loss": 1.7232,
|
| 20548 |
+
"step": 2934
|
| 20549 |
+
},
|
| 20550 |
+
{
|
| 20551 |
+
"epoch": 0.27171208702192906,
|
| 20552 |
+
"grad_norm": 0.14453125,
|
| 20553 |
+
"learning_rate": 0.02,
|
| 20554 |
+
"loss": 1.6562,
|
| 20555 |
+
"step": 2935
|
| 20556 |
+
},
|
| 20557 |
+
{
|
| 20558 |
+
"epoch": 0.2718046635422091,
|
| 20559 |
+
"grad_norm": 0.1552734375,
|
| 20560 |
+
"learning_rate": 0.02,
|
| 20561 |
+
"loss": 1.6476,
|
| 20562 |
+
"step": 2936
|
| 20563 |
+
},
|
| 20564 |
+
{
|
| 20565 |
+
"epoch": 0.27189724006248916,
|
| 20566 |
+
"grad_norm": 0.16015625,
|
| 20567 |
+
"learning_rate": 0.02,
|
| 20568 |
+
"loss": 1.6312,
|
| 20569 |
+
"step": 2937
|
| 20570 |
+
},
|
| 20571 |
+
{
|
| 20572 |
+
"epoch": 0.2719898165827692,
|
| 20573 |
+
"grad_norm": 0.15234375,
|
| 20574 |
+
"learning_rate": 0.02,
|
| 20575 |
+
"loss": 1.7164,
|
| 20576 |
+
"step": 2938
|
| 20577 |
+
},
|
| 20578 |
+
{
|
| 20579 |
+
"epoch": 0.2720823931030492,
|
| 20580 |
+
"grad_norm": 0.1533203125,
|
| 20581 |
+
"learning_rate": 0.02,
|
| 20582 |
+
"loss": 1.6385,
|
| 20583 |
+
"step": 2939
|
| 20584 |
+
},
|
| 20585 |
+
{
|
| 20586 |
+
"epoch": 0.27217496962332927,
|
| 20587 |
+
"grad_norm": 0.1455078125,
|
| 20588 |
+
"learning_rate": 0.02,
|
| 20589 |
+
"loss": 1.6087,
|
| 20590 |
+
"step": 2940
|
| 20591 |
+
},
|
| 20592 |
+
{
|
| 20593 |
+
"epoch": 0.2722675461436093,
|
| 20594 |
+
"grad_norm": 0.1494140625,
|
| 20595 |
+
"learning_rate": 0.02,
|
| 20596 |
+
"loss": 1.6188,
|
| 20597 |
+
"step": 2941
|
| 20598 |
+
},
|
| 20599 |
+
{
|
| 20600 |
+
"epoch": 0.27236012266388937,
|
| 20601 |
+
"grad_norm": 0.142578125,
|
| 20602 |
+
"learning_rate": 0.02,
|
| 20603 |
+
"loss": 1.6092,
|
| 20604 |
+
"step": 2942
|
| 20605 |
+
},
|
| 20606 |
+
{
|
| 20607 |
+
"epoch": 0.2724526991841694,
|
| 20608 |
+
"grad_norm": 0.14453125,
|
| 20609 |
+
"learning_rate": 0.02,
|
| 20610 |
+
"loss": 1.6976,
|
| 20611 |
+
"step": 2943
|
| 20612 |
+
},
|
| 20613 |
+
{
|
| 20614 |
+
"epoch": 0.2725452757044495,
|
| 20615 |
+
"grad_norm": 0.1416015625,
|
| 20616 |
+
"learning_rate": 0.02,
|
| 20617 |
+
"loss": 1.6015,
|
| 20618 |
+
"step": 2944
|
| 20619 |
+
},
|
| 20620 |
+
{
|
| 20621 |
+
"epoch": 0.2726378522247295,
|
| 20622 |
+
"grad_norm": 0.154296875,
|
| 20623 |
+
"learning_rate": 0.02,
|
| 20624 |
+
"loss": 1.6333,
|
| 20625 |
+
"step": 2945
|
| 20626 |
+
},
|
| 20627 |
+
{
|
| 20628 |
+
"epoch": 0.2727304287450095,
|
| 20629 |
+
"grad_norm": 0.158203125,
|
| 20630 |
+
"learning_rate": 0.02,
|
| 20631 |
+
"loss": 1.7133,
|
| 20632 |
+
"step": 2946
|
| 20633 |
+
},
|
| 20634 |
+
{
|
| 20635 |
+
"epoch": 0.2728230052652896,
|
| 20636 |
+
"grad_norm": 0.1533203125,
|
| 20637 |
+
"learning_rate": 0.02,
|
| 20638 |
+
"loss": 1.6359,
|
| 20639 |
+
"step": 2947
|
| 20640 |
+
},
|
| 20641 |
+
{
|
| 20642 |
+
"epoch": 0.2729155817855696,
|
| 20643 |
+
"grad_norm": 0.15625,
|
| 20644 |
+
"learning_rate": 0.02,
|
| 20645 |
+
"loss": 1.5624,
|
| 20646 |
+
"step": 2948
|
| 20647 |
+
},
|
| 20648 |
+
{
|
| 20649 |
+
"epoch": 0.2730081583058497,
|
| 20650 |
+
"grad_norm": 0.154296875,
|
| 20651 |
+
"learning_rate": 0.02,
|
| 20652 |
+
"loss": 1.6915,
|
| 20653 |
+
"step": 2949
|
| 20654 |
+
},
|
| 20655 |
+
{
|
| 20656 |
+
"epoch": 0.27310073482612973,
|
| 20657 |
+
"grad_norm": 0.154296875,
|
| 20658 |
+
"learning_rate": 0.02,
|
| 20659 |
+
"loss": 1.647,
|
| 20660 |
+
"step": 2950
|
| 20661 |
+
},
|
| 20662 |
+
{
|
| 20663 |
+
"epoch": 0.2731933113464098,
|
| 20664 |
+
"grad_norm": 0.158203125,
|
| 20665 |
+
"learning_rate": 0.02,
|
| 20666 |
+
"loss": 1.7084,
|
| 20667 |
+
"step": 2951
|
| 20668 |
+
},
|
| 20669 |
+
{
|
| 20670 |
+
"epoch": 0.2731933113464098,
|
| 20671 |
+
"eval_loss": 1.6467796564102173,
|
| 20672 |
+
"eval_runtime": 63.8421,
|
| 20673 |
+
"eval_samples_per_second": 23.871,
|
| 20674 |
+
"eval_steps_per_second": 5.968,
|
| 20675 |
+
"step": 2951
|
| 20676 |
+
},
|
| 20677 |
+
{
|
| 20678 |
+
"epoch": 0.27328588786668984,
|
| 20679 |
+
"grad_norm": 0.1552734375,
|
| 20680 |
+
"learning_rate": 0.02,
|
| 20681 |
+
"loss": 1.6319,
|
| 20682 |
+
"step": 2952
|
| 20683 |
}
|
| 20684 |
],
|
| 20685 |
"logging_steps": 1,
|
|
|
|
| 20693 |
"should_epoch_stop": false,
|
| 20694 |
"should_evaluate": false,
|
| 20695 |
"should_log": false,
|
| 20696 |
+
"should_save": true,
|
| 20697 |
"should_training_stop": false
|
| 20698 |
},
|
| 20699 |
"attributes": {}
|
| 20700 |
}
|
| 20701 |
},
|
| 20702 |
+
"total_flos": 2.600022217489206e+18,
|
| 20703 |
"train_batch_size": 8,
|
| 20704 |
"trial_name": null,
|
| 20705 |
"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 6289
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:a499b9bb471364e5825828d6af1b283be4f20e4fd8b98f02f170cadf22e490d7
|
| 3 |
size 6289
|