Training in progress, step 10
Browse files- config.json +1 -1
- eval_results_clm.txt +1 -0
- pytorch_model.bin +1 -1
- tokenizer_config.json +1 -1
- train_results.txt +5 -0
- trainer_state.json +1283 -0
- training_args.bin +1 -1
config.json
CHANGED
|
@@ -35,7 +35,7 @@
|
|
| 35 |
}
|
| 36 |
},
|
| 37 |
"torch_dtype": "float32",
|
| 38 |
-
"transformers_version": "4.
|
| 39 |
"use_cache": true,
|
| 40 |
"vocab_size": 50259
|
| 41 |
}
|
|
|
|
| 35 |
}
|
| 36 |
},
|
| 37 |
"torch_dtype": "float32",
|
| 38 |
+
"transformers_version": "4.28.0",
|
| 39 |
"use_cache": true,
|
| 40 |
"vocab_size": 50259
|
| 41 |
}
|
eval_results_clm.txt
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
perplexity = 1.7089595550025904
|
pytorch_model.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1444575065
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:c8631afeff179312fc4aea6b272b587c7ca192a170ae145278abc48782247fbc
|
| 3 |
size 1444575065
|
tokenizer_config.json
CHANGED
|
@@ -1,9 +1,9 @@
|
|
| 1 |
{
|
| 2 |
"add_prefix_space": false,
|
| 3 |
"bos_token": "<|endoftext|>",
|
|
|
|
| 4 |
"eos_token": "<|endoftext|>",
|
| 5 |
"model_max_length": 1024,
|
| 6 |
-
"special_tokens_map_file": null,
|
| 7 |
"tokenizer_class": "GPT2Tokenizer",
|
| 8 |
"unk_token": "<|endoftext|>"
|
| 9 |
}
|
|
|
|
| 1 |
{
|
| 2 |
"add_prefix_space": false,
|
| 3 |
"bos_token": "<|endoftext|>",
|
| 4 |
+
"clean_up_tokenization_spaces": true,
|
| 5 |
"eos_token": "<|endoftext|>",
|
| 6 |
"model_max_length": 1024,
|
|
|
|
| 7 |
"tokenizer_class": "GPT2Tokenizer",
|
| 8 |
"unk_token": "<|endoftext|>"
|
| 9 |
}
|
train_results.txt
ADDED
|
@@ -0,0 +1,5 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
epoch = 0.74
|
| 2 |
+
train_loss = 5.281058074332572
|
| 3 |
+
train_runtime = 1534.4744
|
| 4 |
+
train_samples_per_second = 7.826
|
| 5 |
+
train_steps_per_second = 0.979
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1283 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_metric": 0.5358847379684448,
|
| 3 |
+
"best_model_checkpoint": "/data1/attanasiog/counterspeech/models/gpt2-kl_1_05-hs_cn/checkpoint-340",
|
| 4 |
+
"epoch": 0.7385229540918163,
|
| 5 |
+
"global_step": 370,
|
| 6 |
+
"is_hyper_param_search": false,
|
| 7 |
+
"is_local_process_zero": true,
|
| 8 |
+
"is_world_process_zero": true,
|
| 9 |
+
"log_history": [
|
| 10 |
+
{
|
| 11 |
+
"epoch": 0,
|
| 12 |
+
"kl/loss": 0.40035751461982727,
|
| 13 |
+
"kl/reg_strength": 1.0,
|
| 14 |
+
"kl/scaled_loss": 0.40035751461982727,
|
| 15 |
+
"step": 0
|
| 16 |
+
},
|
| 17 |
+
{
|
| 18 |
+
"epoch": 0.01,
|
| 19 |
+
"learning_rate": 1.6556291390728478e-06,
|
| 20 |
+
"loss": 78.5974,
|
| 21 |
+
"step": 5
|
| 22 |
+
},
|
| 23 |
+
{
|
| 24 |
+
"epoch": 0.01,
|
| 25 |
+
"kl/loss": 0.5881348848342896,
|
| 26 |
+
"kl/reg_strength": 1.0,
|
| 27 |
+
"kl/scaled_loss": 0.5881348848342896,
|
| 28 |
+
"step": 5
|
| 29 |
+
},
|
| 30 |
+
{
|
| 31 |
+
"epoch": 0.02,
|
| 32 |
+
"learning_rate": 3.3112582781456956e-06,
|
| 33 |
+
"loss": 73.9845,
|
| 34 |
+
"step": 10
|
| 35 |
+
},
|
| 36 |
+
{
|
| 37 |
+
"epoch": 0.02,
|
| 38 |
+
"eval_loss": 69.57203674316406,
|
| 39 |
+
"eval_runtime": 4.8116,
|
| 40 |
+
"eval_samples_per_second": 103.915,
|
| 41 |
+
"eval_steps_per_second": 25.979,
|
| 42 |
+
"step": 10
|
| 43 |
+
},
|
| 44 |
+
{
|
| 45 |
+
"epoch": 0.02,
|
| 46 |
+
"kl/loss": 0.6112977862358093,
|
| 47 |
+
"kl/reg_strength": 1.0,
|
| 48 |
+
"kl/scaled_loss": 0.6112977862358093,
|
| 49 |
+
"step": 10
|
| 50 |
+
},
|
| 51 |
+
{
|
| 52 |
+
"epoch": 0.03,
|
| 53 |
+
"learning_rate": 4.966887417218543e-06,
|
| 54 |
+
"loss": 63.0483,
|
| 55 |
+
"step": 15
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.03,
|
| 59 |
+
"kl/loss": 0.502070963382721,
|
| 60 |
+
"kl/reg_strength": 1.0,
|
| 61 |
+
"kl/scaled_loss": 0.502070963382721,
|
| 62 |
+
"step": 15
|
| 63 |
+
},
|
| 64 |
+
{
|
| 65 |
+
"epoch": 0.04,
|
| 66 |
+
"learning_rate": 6.622516556291391e-06,
|
| 67 |
+
"loss": 46.5054,
|
| 68 |
+
"step": 20
|
| 69 |
+
},
|
| 70 |
+
{
|
| 71 |
+
"epoch": 0.04,
|
| 72 |
+
"eval_loss": 32.94132995605469,
|
| 73 |
+
"eval_runtime": 4.8295,
|
| 74 |
+
"eval_samples_per_second": 103.531,
|
| 75 |
+
"eval_steps_per_second": 25.883,
|
| 76 |
+
"step": 20
|
| 77 |
+
},
|
| 78 |
+
{
|
| 79 |
+
"epoch": 0.04,
|
| 80 |
+
"kl/loss": 0.3394828736782074,
|
| 81 |
+
"kl/reg_strength": 1.0,
|
| 82 |
+
"kl/scaled_loss": 0.3394828736782074,
|
| 83 |
+
"step": 20
|
| 84 |
+
},
|
| 85 |
+
{
|
| 86 |
+
"epoch": 0.05,
|
| 87 |
+
"learning_rate": 8.27814569536424e-06,
|
| 88 |
+
"loss": 23.2327,
|
| 89 |
+
"step": 25
|
| 90 |
+
},
|
| 91 |
+
{
|
| 92 |
+
"epoch": 0.05,
|
| 93 |
+
"kl/loss": 0.3881781995296478,
|
| 94 |
+
"kl/reg_strength": 1.0,
|
| 95 |
+
"kl/scaled_loss": 0.3881781995296478,
|
| 96 |
+
"step": 25
|
| 97 |
+
},
|
| 98 |
+
{
|
| 99 |
+
"epoch": 0.06,
|
| 100 |
+
"learning_rate": 9.933774834437086e-06,
|
| 101 |
+
"loss": 13.9547,
|
| 102 |
+
"step": 30
|
| 103 |
+
},
|
| 104 |
+
{
|
| 105 |
+
"epoch": 0.06,
|
| 106 |
+
"eval_loss": 10.638138771057129,
|
| 107 |
+
"eval_runtime": 5.0154,
|
| 108 |
+
"eval_samples_per_second": 99.694,
|
| 109 |
+
"eval_steps_per_second": 24.923,
|
| 110 |
+
"step": 30
|
| 111 |
+
},
|
| 112 |
+
{
|
| 113 |
+
"epoch": 0.06,
|
| 114 |
+
"kl/loss": 0.5543301701545715,
|
| 115 |
+
"kl/reg_strength": 1.0,
|
| 116 |
+
"kl/scaled_loss": 0.5543301701545715,
|
| 117 |
+
"step": 30
|
| 118 |
+
},
|
| 119 |
+
{
|
| 120 |
+
"epoch": 0.07,
|
| 121 |
+
"learning_rate": 1.1589403973509934e-05,
|
| 122 |
+
"loss": 7.8227,
|
| 123 |
+
"step": 35
|
| 124 |
+
},
|
| 125 |
+
{
|
| 126 |
+
"epoch": 0.07,
|
| 127 |
+
"kl/loss": 0.5166069865226746,
|
| 128 |
+
"kl/reg_strength": 1.0,
|
| 129 |
+
"kl/scaled_loss": 0.5166069865226746,
|
| 130 |
+
"step": 35
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.08,
|
| 134 |
+
"learning_rate": 1.3245033112582782e-05,
|
| 135 |
+
"loss": 7.1893,
|
| 136 |
+
"step": 40
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.08,
|
| 140 |
+
"eval_loss": 4.215758800506592,
|
| 141 |
+
"eval_runtime": 4.8956,
|
| 142 |
+
"eval_samples_per_second": 102.133,
|
| 143 |
+
"eval_steps_per_second": 25.533,
|
| 144 |
+
"step": 40
|
| 145 |
+
},
|
| 146 |
+
{
|
| 147 |
+
"epoch": 0.08,
|
| 148 |
+
"kl/loss": 0.49500563740730286,
|
| 149 |
+
"kl/reg_strength": 1.0,
|
| 150 |
+
"kl/scaled_loss": 0.49500563740730286,
|
| 151 |
+
"step": 40
|
| 152 |
+
},
|
| 153 |
+
{
|
| 154 |
+
"epoch": 0.09,
|
| 155 |
+
"learning_rate": 1.490066225165563e-05,
|
| 156 |
+
"loss": 3.8233,
|
| 157 |
+
"step": 45
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.09,
|
| 161 |
+
"kl/loss": 0.4241950511932373,
|
| 162 |
+
"kl/reg_strength": 1.0,
|
| 163 |
+
"kl/scaled_loss": 0.4241950511932373,
|
| 164 |
+
"step": 45
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.1,
|
| 168 |
+
"learning_rate": 1.655629139072848e-05,
|
| 169 |
+
"loss": 4.0234,
|
| 170 |
+
"step": 50
|
| 171 |
+
},
|
| 172 |
+
{
|
| 173 |
+
"epoch": 0.1,
|
| 174 |
+
"eval_loss": 2.051748275756836,
|
| 175 |
+
"eval_runtime": 4.9056,
|
| 176 |
+
"eval_samples_per_second": 101.925,
|
| 177 |
+
"eval_steps_per_second": 25.481,
|
| 178 |
+
"step": 50
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.1,
|
| 182 |
+
"kl/loss": 0.23962484300136566,
|
| 183 |
+
"kl/reg_strength": 1.0,
|
| 184 |
+
"kl/scaled_loss": 0.23962484300136566,
|
| 185 |
+
"step": 50
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.11,
|
| 189 |
+
"learning_rate": 1.8211920529801323e-05,
|
| 190 |
+
"loss": 2.5062,
|
| 191 |
+
"step": 55
|
| 192 |
+
},
|
| 193 |
+
{
|
| 194 |
+
"epoch": 0.11,
|
| 195 |
+
"kl/loss": 0.3255242109298706,
|
| 196 |
+
"kl/reg_strength": 1.0,
|
| 197 |
+
"kl/scaled_loss": 0.3255242109298706,
|
| 198 |
+
"step": 55
|
| 199 |
+
},
|
| 200 |
+
{
|
| 201 |
+
"epoch": 0.12,
|
| 202 |
+
"learning_rate": 1.9867549668874173e-05,
|
| 203 |
+
"loss": 1.8675,
|
| 204 |
+
"step": 60
|
| 205 |
+
},
|
| 206 |
+
{
|
| 207 |
+
"epoch": 0.12,
|
| 208 |
+
"eval_loss": 1.094427227973938,
|
| 209 |
+
"eval_runtime": 4.904,
|
| 210 |
+
"eval_samples_per_second": 101.958,
|
| 211 |
+
"eval_steps_per_second": 25.489,
|
| 212 |
+
"step": 60
|
| 213 |
+
},
|
| 214 |
+
{
|
| 215 |
+
"epoch": 0.12,
|
| 216 |
+
"kl/loss": 0.48757195472717285,
|
| 217 |
+
"kl/reg_strength": 1.0,
|
| 218 |
+
"kl/scaled_loss": 0.48757195472717285,
|
| 219 |
+
"step": 60
|
| 220 |
+
},
|
| 221 |
+
{
|
| 222 |
+
"epoch": 0.13,
|
| 223 |
+
"learning_rate": 2.152317880794702e-05,
|
| 224 |
+
"loss": 1.6018,
|
| 225 |
+
"step": 65
|
| 226 |
+
},
|
| 227 |
+
{
|
| 228 |
+
"epoch": 0.13,
|
| 229 |
+
"kl/loss": 0.4712750017642975,
|
| 230 |
+
"kl/reg_strength": 1.0,
|
| 231 |
+
"kl/scaled_loss": 0.4712750017642975,
|
| 232 |
+
"step": 65
|
| 233 |
+
},
|
| 234 |
+
{
|
| 235 |
+
"epoch": 0.14,
|
| 236 |
+
"learning_rate": 2.317880794701987e-05,
|
| 237 |
+
"loss": 1.4112,
|
| 238 |
+
"step": 70
|
| 239 |
+
},
|
| 240 |
+
{
|
| 241 |
+
"epoch": 0.14,
|
| 242 |
+
"eval_loss": 0.8926725387573242,
|
| 243 |
+
"eval_runtime": 4.8655,
|
| 244 |
+
"eval_samples_per_second": 102.764,
|
| 245 |
+
"eval_steps_per_second": 25.691,
|
| 246 |
+
"step": 70
|
| 247 |
+
},
|
| 248 |
+
{
|
| 249 |
+
"epoch": 0.14,
|
| 250 |
+
"kl/loss": 0.379312127828598,
|
| 251 |
+
"kl/reg_strength": 1.0,
|
| 252 |
+
"kl/scaled_loss": 0.379312127828598,
|
| 253 |
+
"step": 70
|
| 254 |
+
},
|
| 255 |
+
{
|
| 256 |
+
"epoch": 0.15,
|
| 257 |
+
"learning_rate": 2.4834437086092715e-05,
|
| 258 |
+
"loss": 1.1908,
|
| 259 |
+
"step": 75
|
| 260 |
+
},
|
| 261 |
+
{
|
| 262 |
+
"epoch": 0.15,
|
| 263 |
+
"kl/loss": 0.2659462094306946,
|
| 264 |
+
"kl/reg_strength": 1.0,
|
| 265 |
+
"kl/scaled_loss": 0.2659462094306946,
|
| 266 |
+
"step": 75
|
| 267 |
+
},
|
| 268 |
+
{
|
| 269 |
+
"epoch": 0.16,
|
| 270 |
+
"learning_rate": 2.6490066225165565e-05,
|
| 271 |
+
"loss": 1.3181,
|
| 272 |
+
"step": 80
|
| 273 |
+
},
|
| 274 |
+
{
|
| 275 |
+
"epoch": 0.16,
|
| 276 |
+
"eval_loss": 0.8277634978294373,
|
| 277 |
+
"eval_runtime": 4.8909,
|
| 278 |
+
"eval_samples_per_second": 102.231,
|
| 279 |
+
"eval_steps_per_second": 25.558,
|
| 280 |
+
"step": 80
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.16,
|
| 284 |
+
"kl/loss": 0.34173423051834106,
|
| 285 |
+
"kl/reg_strength": 1.0,
|
| 286 |
+
"kl/scaled_loss": 0.34173423051834106,
|
| 287 |
+
"step": 80
|
| 288 |
+
},
|
| 289 |
+
{
|
| 290 |
+
"epoch": 0.17,
|
| 291 |
+
"learning_rate": 2.814569536423841e-05,
|
| 292 |
+
"loss": 1.2674,
|
| 293 |
+
"step": 85
|
| 294 |
+
},
|
| 295 |
+
{
|
| 296 |
+
"epoch": 0.17,
|
| 297 |
+
"kl/loss": 0.49424389004707336,
|
| 298 |
+
"kl/reg_strength": 1.0,
|
| 299 |
+
"kl/scaled_loss": 0.49424389004707336,
|
| 300 |
+
"step": 85
|
| 301 |
+
},
|
| 302 |
+
{
|
| 303 |
+
"epoch": 0.18,
|
| 304 |
+
"learning_rate": 2.980132450331126e-05,
|
| 305 |
+
"loss": 1.1966,
|
| 306 |
+
"step": 90
|
| 307 |
+
},
|
| 308 |
+
{
|
| 309 |
+
"epoch": 0.18,
|
| 310 |
+
"eval_loss": 0.6390898823738098,
|
| 311 |
+
"eval_runtime": 4.8974,
|
| 312 |
+
"eval_samples_per_second": 102.094,
|
| 313 |
+
"eval_steps_per_second": 25.524,
|
| 314 |
+
"step": 90
|
| 315 |
+
},
|
| 316 |
+
{
|
| 317 |
+
"epoch": 0.18,
|
| 318 |
+
"kl/loss": 0.38853219151496887,
|
| 319 |
+
"kl/reg_strength": 1.0,
|
| 320 |
+
"kl/scaled_loss": 0.38853219151496887,
|
| 321 |
+
"step": 90
|
| 322 |
+
},
|
| 323 |
+
{
|
| 324 |
+
"epoch": 0.19,
|
| 325 |
+
"learning_rate": 3.145695364238411e-05,
|
| 326 |
+
"loss": 1.0466,
|
| 327 |
+
"step": 95
|
| 328 |
+
},
|
| 329 |
+
{
|
| 330 |
+
"epoch": 0.19,
|
| 331 |
+
"kl/loss": 0.5124494433403015,
|
| 332 |
+
"kl/reg_strength": 1.0,
|
| 333 |
+
"kl/scaled_loss": 0.5124494433403015,
|
| 334 |
+
"step": 95
|
| 335 |
+
},
|
| 336 |
+
{
|
| 337 |
+
"epoch": 0.2,
|
| 338 |
+
"learning_rate": 3.311258278145696e-05,
|
| 339 |
+
"loss": 1.1266,
|
| 340 |
+
"step": 100
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.2,
|
| 344 |
+
"eval_loss": 0.6523621678352356,
|
| 345 |
+
"eval_runtime": 4.9084,
|
| 346 |
+
"eval_samples_per_second": 101.867,
|
| 347 |
+
"eval_steps_per_second": 25.467,
|
| 348 |
+
"step": 100
|
| 349 |
+
},
|
| 350 |
+
{
|
| 351 |
+
"epoch": 0.2,
|
| 352 |
+
"kl/loss": 0.5430631041526794,
|
| 353 |
+
"kl/reg_strength": 1.0,
|
| 354 |
+
"kl/scaled_loss": 0.5430631041526794,
|
| 355 |
+
"step": 100
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 0.21,
|
| 359 |
+
"learning_rate": 3.47682119205298e-05,
|
| 360 |
+
"loss": 1.0234,
|
| 361 |
+
"step": 105
|
| 362 |
+
},
|
| 363 |
+
{
|
| 364 |
+
"epoch": 0.21,
|
| 365 |
+
"kl/loss": 0.34535568952560425,
|
| 366 |
+
"kl/reg_strength": 1.0,
|
| 367 |
+
"kl/scaled_loss": 0.34535568952560425,
|
| 368 |
+
"step": 105
|
| 369 |
+
},
|
| 370 |
+
{
|
| 371 |
+
"epoch": 0.22,
|
| 372 |
+
"learning_rate": 3.6423841059602646e-05,
|
| 373 |
+
"loss": 1.0109,
|
| 374 |
+
"step": 110
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.22,
|
| 378 |
+
"eval_loss": 0.6015579700469971,
|
| 379 |
+
"eval_runtime": 4.9114,
|
| 380 |
+
"eval_samples_per_second": 101.805,
|
| 381 |
+
"eval_steps_per_second": 25.451,
|
| 382 |
+
"step": 110
|
| 383 |
+
},
|
| 384 |
+
{
|
| 385 |
+
"epoch": 0.22,
|
| 386 |
+
"kl/loss": 0.5453004240989685,
|
| 387 |
+
"kl/reg_strength": 1.0,
|
| 388 |
+
"kl/scaled_loss": 0.5453004240989685,
|
| 389 |
+
"step": 110
|
| 390 |
+
},
|
| 391 |
+
{
|
| 392 |
+
"epoch": 0.23,
|
| 393 |
+
"learning_rate": 3.80794701986755e-05,
|
| 394 |
+
"loss": 1.1632,
|
| 395 |
+
"step": 115
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.23,
|
| 399 |
+
"kl/loss": 0.45745742321014404,
|
| 400 |
+
"kl/reg_strength": 1.0,
|
| 401 |
+
"kl/scaled_loss": 0.45745742321014404,
|
| 402 |
+
"step": 115
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.24,
|
| 406 |
+
"learning_rate": 3.9735099337748346e-05,
|
| 407 |
+
"loss": 1.0176,
|
| 408 |
+
"step": 120
|
| 409 |
+
},
|
| 410 |
+
{
|
| 411 |
+
"epoch": 0.24,
|
| 412 |
+
"eval_loss": 0.5837445259094238,
|
| 413 |
+
"eval_runtime": 4.9098,
|
| 414 |
+
"eval_samples_per_second": 101.836,
|
| 415 |
+
"eval_steps_per_second": 25.459,
|
| 416 |
+
"step": 120
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.24,
|
| 420 |
+
"kl/loss": 0.3508429229259491,
|
| 421 |
+
"kl/reg_strength": 1.0,
|
| 422 |
+
"kl/scaled_loss": 0.3508429229259491,
|
| 423 |
+
"step": 120
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.25,
|
| 427 |
+
"learning_rate": 4.139072847682119e-05,
|
| 428 |
+
"loss": 0.8927,
|
| 429 |
+
"step": 125
|
| 430 |
+
},
|
| 431 |
+
{
|
| 432 |
+
"epoch": 0.25,
|
| 433 |
+
"kl/loss": 0.2840110957622528,
|
| 434 |
+
"kl/reg_strength": 1.0,
|
| 435 |
+
"kl/scaled_loss": 0.2840110957622528,
|
| 436 |
+
"step": 125
|
| 437 |
+
},
|
| 438 |
+
{
|
| 439 |
+
"epoch": 0.26,
|
| 440 |
+
"learning_rate": 4.304635761589404e-05,
|
| 441 |
+
"loss": 1.0522,
|
| 442 |
+
"step": 130
|
| 443 |
+
},
|
| 444 |
+
{
|
| 445 |
+
"epoch": 0.26,
|
| 446 |
+
"eval_loss": 0.5828917622566223,
|
| 447 |
+
"eval_runtime": 4.8679,
|
| 448 |
+
"eval_samples_per_second": 102.713,
|
| 449 |
+
"eval_steps_per_second": 25.678,
|
| 450 |
+
"step": 130
|
| 451 |
+
},
|
| 452 |
+
{
|
| 453 |
+
"epoch": 0.26,
|
| 454 |
+
"kl/loss": 0.3794195055961609,
|
| 455 |
+
"kl/reg_strength": 1.0,
|
| 456 |
+
"kl/scaled_loss": 0.3794195055961609,
|
| 457 |
+
"step": 130
|
| 458 |
+
},
|
| 459 |
+
{
|
| 460 |
+
"epoch": 0.27,
|
| 461 |
+
"learning_rate": 4.470198675496689e-05,
|
| 462 |
+
"loss": 1.0496,
|
| 463 |
+
"step": 135
|
| 464 |
+
},
|
| 465 |
+
{
|
| 466 |
+
"epoch": 0.27,
|
| 467 |
+
"kl/loss": 0.5378769040107727,
|
| 468 |
+
"kl/reg_strength": 1.0,
|
| 469 |
+
"kl/scaled_loss": 0.5378769040107727,
|
| 470 |
+
"step": 135
|
| 471 |
+
},
|
| 472 |
+
{
|
| 473 |
+
"epoch": 0.28,
|
| 474 |
+
"learning_rate": 4.635761589403974e-05,
|
| 475 |
+
"loss": 1.0986,
|
| 476 |
+
"step": 140
|
| 477 |
+
},
|
| 478 |
+
{
|
| 479 |
+
"epoch": 0.28,
|
| 480 |
+
"eval_loss": 0.5824211239814758,
|
| 481 |
+
"eval_runtime": 4.8869,
|
| 482 |
+
"eval_samples_per_second": 102.315,
|
| 483 |
+
"eval_steps_per_second": 25.579,
|
| 484 |
+
"step": 140
|
| 485 |
+
},
|
| 486 |
+
{
|
| 487 |
+
"epoch": 0.28,
|
| 488 |
+
"kl/loss": 0.37985092401504517,
|
| 489 |
+
"kl/reg_strength": 1.0,
|
| 490 |
+
"kl/scaled_loss": 0.37985092401504517,
|
| 491 |
+
"step": 140
|
| 492 |
+
},
|
| 493 |
+
{
|
| 494 |
+
"epoch": 0.29,
|
| 495 |
+
"learning_rate": 4.8013245033112584e-05,
|
| 496 |
+
"loss": 0.9009,
|
| 497 |
+
"step": 145
|
| 498 |
+
},
|
| 499 |
+
{
|
| 500 |
+
"epoch": 0.29,
|
| 501 |
+
"kl/loss": 0.46719425916671753,
|
| 502 |
+
"kl/reg_strength": 1.0,
|
| 503 |
+
"kl/scaled_loss": 0.46719425916671753,
|
| 504 |
+
"step": 145
|
| 505 |
+
},
|
| 506 |
+
{
|
| 507 |
+
"epoch": 0.3,
|
| 508 |
+
"learning_rate": 4.966887417218543e-05,
|
| 509 |
+
"loss": 1.1256,
|
| 510 |
+
"step": 150
|
| 511 |
+
},
|
| 512 |
+
{
|
| 513 |
+
"epoch": 0.3,
|
| 514 |
+
"eval_loss": 0.5937540531158447,
|
| 515 |
+
"eval_runtime": 4.9008,
|
| 516 |
+
"eval_samples_per_second": 102.024,
|
| 517 |
+
"eval_steps_per_second": 25.506,
|
| 518 |
+
"step": 150
|
| 519 |
+
},
|
| 520 |
+
{
|
| 521 |
+
"epoch": 0.3,
|
| 522 |
+
"kl/loss": 0.5251249074935913,
|
| 523 |
+
"kl/reg_strength": 1.0,
|
| 524 |
+
"kl/scaled_loss": 0.5251249074935913,
|
| 525 |
+
"step": 150
|
| 526 |
+
},
|
| 527 |
+
{
|
| 528 |
+
"epoch": 0.31,
|
| 529 |
+
"learning_rate": 4.985207100591716e-05,
|
| 530 |
+
"loss": 1.0179,
|
| 531 |
+
"step": 155
|
| 532 |
+
},
|
| 533 |
+
{
|
| 534 |
+
"epoch": 0.31,
|
| 535 |
+
"kl/loss": 0.4942081570625305,
|
| 536 |
+
"kl/reg_strength": 1.0,
|
| 537 |
+
"kl/scaled_loss": 0.4942081570625305,
|
| 538 |
+
"step": 155
|
| 539 |
+
},
|
| 540 |
+
{
|
| 541 |
+
"epoch": 0.32,
|
| 542 |
+
"learning_rate": 4.966715976331361e-05,
|
| 543 |
+
"loss": 1.072,
|
| 544 |
+
"step": 160
|
| 545 |
+
},
|
| 546 |
+
{
|
| 547 |
+
"epoch": 0.32,
|
| 548 |
+
"eval_loss": 0.584824800491333,
|
| 549 |
+
"eval_runtime": 4.904,
|
| 550 |
+
"eval_samples_per_second": 101.958,
|
| 551 |
+
"eval_steps_per_second": 25.489,
|
| 552 |
+
"step": 160
|
| 553 |
+
},
|
| 554 |
+
{
|
| 555 |
+
"epoch": 0.32,
|
| 556 |
+
"kl/loss": 0.3913673460483551,
|
| 557 |
+
"kl/reg_strength": 1.0,
|
| 558 |
+
"kl/scaled_loss": 0.3913673460483551,
|
| 559 |
+
"step": 160
|
| 560 |
+
},
|
| 561 |
+
{
|
| 562 |
+
"epoch": 0.33,
|
| 563 |
+
"learning_rate": 4.948224852071006e-05,
|
| 564 |
+
"loss": 0.9444,
|
| 565 |
+
"step": 165
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.33,
|
| 569 |
+
"kl/loss": 0.47482964396476746,
|
| 570 |
+
"kl/reg_strength": 1.0,
|
| 571 |
+
"kl/scaled_loss": 0.47482964396476746,
|
| 572 |
+
"step": 165
|
| 573 |
+
},
|
| 574 |
+
{
|
| 575 |
+
"epoch": 0.34,
|
| 576 |
+
"learning_rate": 4.929733727810651e-05,
|
| 577 |
+
"loss": 0.9252,
|
| 578 |
+
"step": 170
|
| 579 |
+
},
|
| 580 |
+
{
|
| 581 |
+
"epoch": 0.34,
|
| 582 |
+
"eval_loss": 0.5677751302719116,
|
| 583 |
+
"eval_runtime": 4.9133,
|
| 584 |
+
"eval_samples_per_second": 101.764,
|
| 585 |
+
"eval_steps_per_second": 25.441,
|
| 586 |
+
"step": 170
|
| 587 |
+
},
|
| 588 |
+
{
|
| 589 |
+
"epoch": 0.34,
|
| 590 |
+
"kl/loss": 0.3093286156654358,
|
| 591 |
+
"kl/reg_strength": 1.0,
|
| 592 |
+
"kl/scaled_loss": 0.3093286156654358,
|
| 593 |
+
"step": 170
|
| 594 |
+
},
|
| 595 |
+
{
|
| 596 |
+
"epoch": 0.35,
|
| 597 |
+
"learning_rate": 4.9112426035502965e-05,
|
| 598 |
+
"loss": 0.9672,
|
| 599 |
+
"step": 175
|
| 600 |
+
},
|
| 601 |
+
{
|
| 602 |
+
"epoch": 0.35,
|
| 603 |
+
"kl/loss": 0.5472012758255005,
|
| 604 |
+
"kl/reg_strength": 1.0,
|
| 605 |
+
"kl/scaled_loss": 0.5472012758255005,
|
| 606 |
+
"step": 175
|
| 607 |
+
},
|
| 608 |
+
{
|
| 609 |
+
"epoch": 0.36,
|
| 610 |
+
"learning_rate": 4.892751479289941e-05,
|
| 611 |
+
"loss": 1.0958,
|
| 612 |
+
"step": 180
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.36,
|
| 616 |
+
"eval_loss": 0.560631513595581,
|
| 617 |
+
"eval_runtime": 4.9131,
|
| 618 |
+
"eval_samples_per_second": 101.769,
|
| 619 |
+
"eval_steps_per_second": 25.442,
|
| 620 |
+
"step": 180
|
| 621 |
+
},
|
| 622 |
+
{
|
| 623 |
+
"epoch": 0.36,
|
| 624 |
+
"kl/loss": 0.3398216664791107,
|
| 625 |
+
"kl/reg_strength": 1.0,
|
| 626 |
+
"kl/scaled_loss": 0.3398216664791107,
|
| 627 |
+
"step": 180
|
| 628 |
+
},
|
| 629 |
+
{
|
| 630 |
+
"epoch": 0.37,
|
| 631 |
+
"learning_rate": 4.8742603550295864e-05,
|
| 632 |
+
"loss": 1.0966,
|
| 633 |
+
"step": 185
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.37,
|
| 637 |
+
"kl/loss": 0.5004676580429077,
|
| 638 |
+
"kl/reg_strength": 1.0,
|
| 639 |
+
"kl/scaled_loss": 0.5004676580429077,
|
| 640 |
+
"step": 185
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.38,
|
| 644 |
+
"learning_rate": 4.855769230769231e-05,
|
| 645 |
+
"loss": 1.1286,
|
| 646 |
+
"step": 190
|
| 647 |
+
},
|
| 648 |
+
{
|
| 649 |
+
"epoch": 0.38,
|
| 650 |
+
"eval_loss": 0.5572844743728638,
|
| 651 |
+
"eval_runtime": 4.8589,
|
| 652 |
+
"eval_samples_per_second": 102.904,
|
| 653 |
+
"eval_steps_per_second": 25.726,
|
| 654 |
+
"step": 190
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.38,
|
| 658 |
+
"kl/loss": 0.43443694710731506,
|
| 659 |
+
"kl/reg_strength": 1.0,
|
| 660 |
+
"kl/scaled_loss": 0.43443694710731506,
|
| 661 |
+
"step": 190
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.39,
|
| 665 |
+
"learning_rate": 4.8372781065088756e-05,
|
| 666 |
+
"loss": 1.0002,
|
| 667 |
+
"step": 195
|
| 668 |
+
},
|
| 669 |
+
{
|
| 670 |
+
"epoch": 0.39,
|
| 671 |
+
"kl/loss": 0.3998401165008545,
|
| 672 |
+
"kl/reg_strength": 1.0,
|
| 673 |
+
"kl/scaled_loss": 0.3998401165008545,
|
| 674 |
+
"step": 195
|
| 675 |
+
},
|
| 676 |
+
{
|
| 677 |
+
"epoch": 0.4,
|
| 678 |
+
"learning_rate": 4.818786982248521e-05,
|
| 679 |
+
"loss": 0.9755,
|
| 680 |
+
"step": 200
|
| 681 |
+
},
|
| 682 |
+
{
|
| 683 |
+
"epoch": 0.4,
|
| 684 |
+
"eval_loss": 0.5549196004867554,
|
| 685 |
+
"eval_runtime": 4.8876,
|
| 686 |
+
"eval_samples_per_second": 102.3,
|
| 687 |
+
"eval_steps_per_second": 25.575,
|
| 688 |
+
"step": 200
|
| 689 |
+
},
|
| 690 |
+
{
|
| 691 |
+
"epoch": 0.4,
|
| 692 |
+
"kl/loss": 0.3528302013874054,
|
| 693 |
+
"kl/reg_strength": 1.0,
|
| 694 |
+
"kl/scaled_loss": 0.3528302013874054,
|
| 695 |
+
"step": 200
|
| 696 |
+
},
|
| 697 |
+
{
|
| 698 |
+
"epoch": 0.41,
|
| 699 |
+
"learning_rate": 4.8002958579881655e-05,
|
| 700 |
+
"loss": 0.955,
|
| 701 |
+
"step": 205
|
| 702 |
+
},
|
| 703 |
+
{
|
| 704 |
+
"epoch": 0.41,
|
| 705 |
+
"kl/loss": 0.5008842945098877,
|
| 706 |
+
"kl/reg_strength": 1.0,
|
| 707 |
+
"kl/scaled_loss": 0.5008842945098877,
|
| 708 |
+
"step": 205
|
| 709 |
+
},
|
| 710 |
+
{
|
| 711 |
+
"epoch": 0.42,
|
| 712 |
+
"learning_rate": 4.781804733727811e-05,
|
| 713 |
+
"loss": 1.1749,
|
| 714 |
+
"step": 210
|
| 715 |
+
},
|
| 716 |
+
{
|
| 717 |
+
"epoch": 0.42,
|
| 718 |
+
"eval_loss": 0.5554329752922058,
|
| 719 |
+
"eval_runtime": 4.9051,
|
| 720 |
+
"eval_samples_per_second": 101.934,
|
| 721 |
+
"eval_steps_per_second": 25.483,
|
| 722 |
+
"step": 210
|
| 723 |
+
},
|
| 724 |
+
{
|
| 725 |
+
"epoch": 0.42,
|
| 726 |
+
"kl/loss": 0.42498648166656494,
|
| 727 |
+
"kl/reg_strength": 1.0,
|
| 728 |
+
"kl/scaled_loss": 0.42498648166656494,
|
| 729 |
+
"step": 210
|
| 730 |
+
},
|
| 731 |
+
{
|
| 732 |
+
"epoch": 0.43,
|
| 733 |
+
"learning_rate": 4.7633136094674555e-05,
|
| 734 |
+
"loss": 1.0792,
|
| 735 |
+
"step": 215
|
| 736 |
+
},
|
| 737 |
+
{
|
| 738 |
+
"epoch": 0.43,
|
| 739 |
+
"kl/loss": 0.6202272176742554,
|
| 740 |
+
"kl/reg_strength": 1.0,
|
| 741 |
+
"kl/scaled_loss": 0.6202272176742554,
|
| 742 |
+
"step": 215
|
| 743 |
+
},
|
| 744 |
+
{
|
| 745 |
+
"epoch": 0.44,
|
| 746 |
+
"learning_rate": 4.744822485207101e-05,
|
| 747 |
+
"loss": 1.1154,
|
| 748 |
+
"step": 220
|
| 749 |
+
},
|
| 750 |
+
{
|
| 751 |
+
"epoch": 0.44,
|
| 752 |
+
"eval_loss": 0.5472630858421326,
|
| 753 |
+
"eval_runtime": 4.9056,
|
| 754 |
+
"eval_samples_per_second": 101.925,
|
| 755 |
+
"eval_steps_per_second": 25.481,
|
| 756 |
+
"step": 220
|
| 757 |
+
},
|
| 758 |
+
{
|
| 759 |
+
"epoch": 0.44,
|
| 760 |
+
"kl/loss": 0.2877408266067505,
|
| 761 |
+
"kl/reg_strength": 1.0,
|
| 762 |
+
"kl/scaled_loss": 0.2877408266067505,
|
| 763 |
+
"step": 220
|
| 764 |
+
},
|
| 765 |
+
{
|
| 766 |
+
"epoch": 0.45,
|
| 767 |
+
"learning_rate": 4.726331360946746e-05,
|
| 768 |
+
"loss": 0.8598,
|
| 769 |
+
"step": 225
|
| 770 |
+
},
|
| 771 |
+
{
|
| 772 |
+
"epoch": 0.45,
|
| 773 |
+
"kl/loss": 0.37519893050193787,
|
| 774 |
+
"kl/reg_strength": 1.0,
|
| 775 |
+
"kl/scaled_loss": 0.37519893050193787,
|
| 776 |
+
"step": 225
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 0.46,
|
| 780 |
+
"learning_rate": 4.707840236686391e-05,
|
| 781 |
+
"loss": 0.9256,
|
| 782 |
+
"step": 230
|
| 783 |
+
},
|
| 784 |
+
{
|
| 785 |
+
"epoch": 0.46,
|
| 786 |
+
"eval_loss": 0.5486032366752625,
|
| 787 |
+
"eval_runtime": 4.9132,
|
| 788 |
+
"eval_samples_per_second": 101.766,
|
| 789 |
+
"eval_steps_per_second": 25.441,
|
| 790 |
+
"step": 230
|
| 791 |
+
},
|
| 792 |
+
{
|
| 793 |
+
"epoch": 0.46,
|
| 794 |
+
"kl/loss": 0.39483198523521423,
|
| 795 |
+
"kl/reg_strength": 1.0,
|
| 796 |
+
"kl/scaled_loss": 0.39483198523521423,
|
| 797 |
+
"step": 230
|
| 798 |
+
},
|
| 799 |
+
{
|
| 800 |
+
"epoch": 0.47,
|
| 801 |
+
"learning_rate": 4.689349112426036e-05,
|
| 802 |
+
"loss": 1.0932,
|
| 803 |
+
"step": 235
|
| 804 |
+
},
|
| 805 |
+
{
|
| 806 |
+
"epoch": 0.47,
|
| 807 |
+
"kl/loss": 0.4944440424442291,
|
| 808 |
+
"kl/reg_strength": 1.0,
|
| 809 |
+
"kl/scaled_loss": 0.4944440424442291,
|
| 810 |
+
"step": 235
|
| 811 |
+
},
|
| 812 |
+
{
|
| 813 |
+
"epoch": 0.48,
|
| 814 |
+
"learning_rate": 4.6708579881656806e-05,
|
| 815 |
+
"loss": 1.006,
|
| 816 |
+
"step": 240
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.48,
|
| 820 |
+
"eval_loss": 0.5463646650314331,
|
| 821 |
+
"eval_runtime": 4.9139,
|
| 822 |
+
"eval_samples_per_second": 101.752,
|
| 823 |
+
"eval_steps_per_second": 25.438,
|
| 824 |
+
"step": 240
|
| 825 |
+
},
|
| 826 |
+
{
|
| 827 |
+
"epoch": 0.48,
|
| 828 |
+
"kl/loss": 0.4611687660217285,
|
| 829 |
+
"kl/reg_strength": 1.0,
|
| 830 |
+
"kl/scaled_loss": 0.4611687660217285,
|
| 831 |
+
"step": 240
|
| 832 |
+
},
|
| 833 |
+
{
|
| 834 |
+
"epoch": 0.49,
|
| 835 |
+
"learning_rate": 4.652366863905326e-05,
|
| 836 |
+
"loss": 0.8044,
|
| 837 |
+
"step": 245
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.49,
|
| 841 |
+
"kl/loss": 0.21632161736488342,
|
| 842 |
+
"kl/reg_strength": 1.0,
|
| 843 |
+
"kl/scaled_loss": 0.21632161736488342,
|
| 844 |
+
"step": 245
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.5,
|
| 848 |
+
"learning_rate": 4.6338757396449705e-05,
|
| 849 |
+
"loss": 0.9398,
|
| 850 |
+
"step": 250
|
| 851 |
+
},
|
| 852 |
+
{
|
| 853 |
+
"epoch": 0.5,
|
| 854 |
+
"eval_loss": 0.5485103130340576,
|
| 855 |
+
"eval_runtime": 4.8683,
|
| 856 |
+
"eval_samples_per_second": 102.706,
|
| 857 |
+
"eval_steps_per_second": 25.677,
|
| 858 |
+
"step": 250
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.5,
|
| 862 |
+
"kl/loss": 0.4572712481021881,
|
| 863 |
+
"kl/reg_strength": 1.0,
|
| 864 |
+
"kl/scaled_loss": 0.4572712481021881,
|
| 865 |
+
"step": 250
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.51,
|
| 869 |
+
"learning_rate": 4.615384615384616e-05,
|
| 870 |
+
"loss": 0.8885,
|
| 871 |
+
"step": 255
|
| 872 |
+
},
|
| 873 |
+
{
|
| 874 |
+
"epoch": 0.51,
|
| 875 |
+
"kl/loss": 0.3353426158428192,
|
| 876 |
+
"kl/reg_strength": 1.0,
|
| 877 |
+
"kl/scaled_loss": 0.3353426158428192,
|
| 878 |
+
"step": 255
|
| 879 |
+
},
|
| 880 |
+
{
|
| 881 |
+
"epoch": 0.52,
|
| 882 |
+
"learning_rate": 4.5968934911242604e-05,
|
| 883 |
+
"loss": 0.9877,
|
| 884 |
+
"step": 260
|
| 885 |
+
},
|
| 886 |
+
{
|
| 887 |
+
"epoch": 0.52,
|
| 888 |
+
"eval_loss": 0.5477557182312012,
|
| 889 |
+
"eval_runtime": 4.8902,
|
| 890 |
+
"eval_samples_per_second": 102.245,
|
| 891 |
+
"eval_steps_per_second": 25.561,
|
| 892 |
+
"step": 260
|
| 893 |
+
},
|
| 894 |
+
{
|
| 895 |
+
"epoch": 0.52,
|
| 896 |
+
"kl/loss": 0.478200763463974,
|
| 897 |
+
"kl/reg_strength": 1.0,
|
| 898 |
+
"kl/scaled_loss": 0.478200763463974,
|
| 899 |
+
"step": 260
|
| 900 |
+
},
|
| 901 |
+
{
|
| 902 |
+
"epoch": 0.53,
|
| 903 |
+
"learning_rate": 4.578402366863905e-05,
|
| 904 |
+
"loss": 0.9594,
|
| 905 |
+
"step": 265
|
| 906 |
+
},
|
| 907 |
+
{
|
| 908 |
+
"epoch": 0.53,
|
| 909 |
+
"kl/loss": 0.4592299163341522,
|
| 910 |
+
"kl/reg_strength": 1.0,
|
| 911 |
+
"kl/scaled_loss": 0.4592299163341522,
|
| 912 |
+
"step": 265
|
| 913 |
+
},
|
| 914 |
+
{
|
| 915 |
+
"epoch": 0.54,
|
| 916 |
+
"learning_rate": 4.5599112426035504e-05,
|
| 917 |
+
"loss": 0.9719,
|
| 918 |
+
"step": 270
|
| 919 |
+
},
|
| 920 |
+
{
|
| 921 |
+
"epoch": 0.54,
|
| 922 |
+
"eval_loss": 0.5442342758178711,
|
| 923 |
+
"eval_runtime": 4.9068,
|
| 924 |
+
"eval_samples_per_second": 101.899,
|
| 925 |
+
"eval_steps_per_second": 25.475,
|
| 926 |
+
"step": 270
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 0.54,
|
| 930 |
+
"kl/loss": 0.5783120393753052,
|
| 931 |
+
"kl/reg_strength": 1.0,
|
| 932 |
+
"kl/scaled_loss": 0.5783120393753052,
|
| 933 |
+
"step": 270
|
| 934 |
+
},
|
| 935 |
+
{
|
| 936 |
+
"epoch": 0.55,
|
| 937 |
+
"learning_rate": 4.5414201183431957e-05,
|
| 938 |
+
"loss": 1.0057,
|
| 939 |
+
"step": 275
|
| 940 |
+
},
|
| 941 |
+
{
|
| 942 |
+
"epoch": 0.55,
|
| 943 |
+
"kl/loss": 0.49252939224243164,
|
| 944 |
+
"kl/reg_strength": 1.0,
|
| 945 |
+
"kl/scaled_loss": 0.49252939224243164,
|
| 946 |
+
"step": 275
|
| 947 |
+
},
|
| 948 |
+
{
|
| 949 |
+
"epoch": 0.56,
|
| 950 |
+
"learning_rate": 4.522928994082841e-05,
|
| 951 |
+
"loss": 1.0415,
|
| 952 |
+
"step": 280
|
| 953 |
+
},
|
| 954 |
+
{
|
| 955 |
+
"epoch": 0.56,
|
| 956 |
+
"eval_loss": 0.5440837144851685,
|
| 957 |
+
"eval_runtime": 4.9086,
|
| 958 |
+
"eval_samples_per_second": 101.862,
|
| 959 |
+
"eval_steps_per_second": 25.465,
|
| 960 |
+
"step": 280
|
| 961 |
+
},
|
| 962 |
+
{
|
| 963 |
+
"epoch": 0.56,
|
| 964 |
+
"kl/loss": 0.38020560145378113,
|
| 965 |
+
"kl/reg_strength": 1.0,
|
| 966 |
+
"kl/scaled_loss": 0.38020560145378113,
|
| 967 |
+
"step": 280
|
| 968 |
+
},
|
| 969 |
+
{
|
| 970 |
+
"epoch": 0.57,
|
| 971 |
+
"learning_rate": 4.5044378698224856e-05,
|
| 972 |
+
"loss": 0.9769,
|
| 973 |
+
"step": 285
|
| 974 |
+
},
|
| 975 |
+
{
|
| 976 |
+
"epoch": 0.57,
|
| 977 |
+
"kl/loss": 0.37340953946113586,
|
| 978 |
+
"kl/reg_strength": 1.0,
|
| 979 |
+
"kl/scaled_loss": 0.37340953946113586,
|
| 980 |
+
"step": 285
|
| 981 |
+
},
|
| 982 |
+
{
|
| 983 |
+
"epoch": 0.58,
|
| 984 |
+
"learning_rate": 4.48594674556213e-05,
|
| 985 |
+
"loss": 0.9641,
|
| 986 |
+
"step": 290
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 0.58,
|
| 990 |
+
"eval_loss": 0.5448547601699829,
|
| 991 |
+
"eval_runtime": 4.9072,
|
| 992 |
+
"eval_samples_per_second": 101.892,
|
| 993 |
+
"eval_steps_per_second": 25.473,
|
| 994 |
+
"step": 290
|
| 995 |
+
},
|
| 996 |
+
{
|
| 997 |
+
"epoch": 0.58,
|
| 998 |
+
"kl/loss": 0.511296272277832,
|
| 999 |
+
"kl/reg_strength": 1.0,
|
| 1000 |
+
"kl/scaled_loss": 0.511296272277832,
|
| 1001 |
+
"step": 290
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 0.59,
|
| 1005 |
+
"learning_rate": 4.4674556213017755e-05,
|
| 1006 |
+
"loss": 1.0768,
|
| 1007 |
+
"step": 295
|
| 1008 |
+
},
|
| 1009 |
+
{
|
| 1010 |
+
"epoch": 0.59,
|
| 1011 |
+
"kl/loss": 0.32476699352264404,
|
| 1012 |
+
"kl/reg_strength": 1.0,
|
| 1013 |
+
"kl/scaled_loss": 0.32476699352264404,
|
| 1014 |
+
"step": 295
|
| 1015 |
+
},
|
| 1016 |
+
{
|
| 1017 |
+
"epoch": 0.6,
|
| 1018 |
+
"learning_rate": 4.44896449704142e-05,
|
| 1019 |
+
"loss": 0.8955,
|
| 1020 |
+
"step": 300
|
| 1021 |
+
},
|
| 1022 |
+
{
|
| 1023 |
+
"epoch": 0.6,
|
| 1024 |
+
"eval_loss": 0.5461843013763428,
|
| 1025 |
+
"eval_runtime": 4.9095,
|
| 1026 |
+
"eval_samples_per_second": 101.844,
|
| 1027 |
+
"eval_steps_per_second": 25.461,
|
| 1028 |
+
"step": 300
|
| 1029 |
+
},
|
| 1030 |
+
{
|
| 1031 |
+
"epoch": 0.6,
|
| 1032 |
+
"kl/loss": 0.4175699055194855,
|
| 1033 |
+
"kl/reg_strength": 1.0,
|
| 1034 |
+
"kl/scaled_loss": 0.4175699055194855,
|
| 1035 |
+
"step": 300
|
| 1036 |
+
},
|
| 1037 |
+
{
|
| 1038 |
+
"epoch": 0.61,
|
| 1039 |
+
"learning_rate": 4.4304733727810654e-05,
|
| 1040 |
+
"loss": 0.9859,
|
| 1041 |
+
"step": 305
|
| 1042 |
+
},
|
| 1043 |
+
{
|
| 1044 |
+
"epoch": 0.61,
|
| 1045 |
+
"kl/loss": 0.40970510244369507,
|
| 1046 |
+
"kl/reg_strength": 1.0,
|
| 1047 |
+
"kl/scaled_loss": 0.40970510244369507,
|
| 1048 |
+
"step": 305
|
| 1049 |
+
},
|
| 1050 |
+
{
|
| 1051 |
+
"epoch": 0.62,
|
| 1052 |
+
"learning_rate": 4.41198224852071e-05,
|
| 1053 |
+
"loss": 1.0203,
|
| 1054 |
+
"step": 310
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.62,
|
| 1058 |
+
"eval_loss": 0.5415293574333191,
|
| 1059 |
+
"eval_runtime": 4.8684,
|
| 1060 |
+
"eval_samples_per_second": 102.702,
|
| 1061 |
+
"eval_steps_per_second": 25.676,
|
| 1062 |
+
"step": 310
|
| 1063 |
+
},
|
| 1064 |
+
{
|
| 1065 |
+
"epoch": 0.62,
|
| 1066 |
+
"kl/loss": 0.5084818005561829,
|
| 1067 |
+
"kl/reg_strength": 1.0,
|
| 1068 |
+
"kl/scaled_loss": 0.5084818005561829,
|
| 1069 |
+
"step": 310
|
| 1070 |
+
},
|
| 1071 |
+
{
|
| 1072 |
+
"epoch": 0.63,
|
| 1073 |
+
"learning_rate": 4.393491124260355e-05,
|
| 1074 |
+
"loss": 1.0563,
|
| 1075 |
+
"step": 315
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 0.63,
|
| 1079 |
+
"kl/loss": 0.3006637394428253,
|
| 1080 |
+
"kl/reg_strength": 1.0,
|
| 1081 |
+
"kl/scaled_loss": 0.3006637394428253,
|
| 1082 |
+
"step": 315
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.64,
|
| 1086 |
+
"learning_rate": 4.375e-05,
|
| 1087 |
+
"loss": 1.1277,
|
| 1088 |
+
"step": 320
|
| 1089 |
+
},
|
| 1090 |
+
{
|
| 1091 |
+
"epoch": 0.64,
|
| 1092 |
+
"eval_loss": 0.5415114164352417,
|
| 1093 |
+
"eval_runtime": 4.8896,
|
| 1094 |
+
"eval_samples_per_second": 102.258,
|
| 1095 |
+
"eval_steps_per_second": 25.565,
|
| 1096 |
+
"step": 320
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 0.64,
|
| 1100 |
+
"kl/loss": 0.4050839841365814,
|
| 1101 |
+
"kl/reg_strength": 1.0,
|
| 1102 |
+
"kl/scaled_loss": 0.4050839841365814,
|
| 1103 |
+
"step": 320
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 0.65,
|
| 1107 |
+
"learning_rate": 4.356508875739645e-05,
|
| 1108 |
+
"loss": 0.8973,
|
| 1109 |
+
"step": 325
|
| 1110 |
+
},
|
| 1111 |
+
{
|
| 1112 |
+
"epoch": 0.65,
|
| 1113 |
+
"kl/loss": 0.49359455704689026,
|
| 1114 |
+
"kl/reg_strength": 1.0,
|
| 1115 |
+
"kl/scaled_loss": 0.49359455704689026,
|
| 1116 |
+
"step": 325
|
| 1117 |
+
},
|
| 1118 |
+
{
|
| 1119 |
+
"epoch": 0.66,
|
| 1120 |
+
"learning_rate": 4.3380177514792906e-05,
|
| 1121 |
+
"loss": 0.943,
|
| 1122 |
+
"step": 330
|
| 1123 |
+
},
|
| 1124 |
+
{
|
| 1125 |
+
"epoch": 0.66,
|
| 1126 |
+
"eval_loss": 0.5408716201782227,
|
| 1127 |
+
"eval_runtime": 4.9035,
|
| 1128 |
+
"eval_samples_per_second": 101.968,
|
| 1129 |
+
"eval_steps_per_second": 25.492,
|
| 1130 |
+
"step": 330
|
| 1131 |
+
},
|
| 1132 |
+
{
|
| 1133 |
+
"epoch": 0.66,
|
| 1134 |
+
"kl/loss": 0.23989157378673553,
|
| 1135 |
+
"kl/reg_strength": 1.0,
|
| 1136 |
+
"kl/scaled_loss": 0.23989157378673553,
|
| 1137 |
+
"step": 330
|
| 1138 |
+
},
|
| 1139 |
+
{
|
| 1140 |
+
"epoch": 0.67,
|
| 1141 |
+
"learning_rate": 4.319526627218935e-05,
|
| 1142 |
+
"loss": 0.9777,
|
| 1143 |
+
"step": 335
|
| 1144 |
+
},
|
| 1145 |
+
{
|
| 1146 |
+
"epoch": 0.67,
|
| 1147 |
+
"kl/loss": 0.49692365527153015,
|
| 1148 |
+
"kl/reg_strength": 1.0,
|
| 1149 |
+
"kl/scaled_loss": 0.49692365527153015,
|
| 1150 |
+
"step": 335
|
| 1151 |
+
},
|
| 1152 |
+
{
|
| 1153 |
+
"epoch": 0.68,
|
| 1154 |
+
"learning_rate": 4.3010355029585805e-05,
|
| 1155 |
+
"loss": 0.9258,
|
| 1156 |
+
"step": 340
|
| 1157 |
+
},
|
| 1158 |
+
{
|
| 1159 |
+
"epoch": 0.68,
|
| 1160 |
+
"eval_loss": 0.5358847379684448,
|
| 1161 |
+
"eval_runtime": 4.91,
|
| 1162 |
+
"eval_samples_per_second": 101.832,
|
| 1163 |
+
"eval_steps_per_second": 25.458,
|
| 1164 |
+
"step": 340
|
| 1165 |
+
},
|
| 1166 |
+
{
|
| 1167 |
+
"epoch": 0.68,
|
| 1168 |
+
"kl/loss": 0.4260445237159729,
|
| 1169 |
+
"kl/reg_strength": 1.0,
|
| 1170 |
+
"kl/scaled_loss": 0.4260445237159729,
|
| 1171 |
+
"step": 340
|
| 1172 |
+
},
|
| 1173 |
+
{
|
| 1174 |
+
"epoch": 0.69,
|
| 1175 |
+
"learning_rate": 4.282544378698225e-05,
|
| 1176 |
+
"loss": 0.9521,
|
| 1177 |
+
"step": 345
|
| 1178 |
+
},
|
| 1179 |
+
{
|
| 1180 |
+
"epoch": 0.69,
|
| 1181 |
+
"kl/loss": 0.4256747364997864,
|
| 1182 |
+
"kl/reg_strength": 1.0,
|
| 1183 |
+
"kl/scaled_loss": 0.4256747364997864,
|
| 1184 |
+
"step": 345
|
| 1185 |
+
},
|
| 1186 |
+
{
|
| 1187 |
+
"epoch": 0.7,
|
| 1188 |
+
"learning_rate": 4.2640532544378704e-05,
|
| 1189 |
+
"loss": 0.9217,
|
| 1190 |
+
"step": 350
|
| 1191 |
+
},
|
| 1192 |
+
{
|
| 1193 |
+
"epoch": 0.7,
|
| 1194 |
+
"eval_loss": 0.5400574803352356,
|
| 1195 |
+
"eval_runtime": 4.9107,
|
| 1196 |
+
"eval_samples_per_second": 101.818,
|
| 1197 |
+
"eval_steps_per_second": 25.454,
|
| 1198 |
+
"step": 350
|
| 1199 |
+
},
|
| 1200 |
+
{
|
| 1201 |
+
"epoch": 0.7,
|
| 1202 |
+
"kl/loss": 0.5429372787475586,
|
| 1203 |
+
"kl/reg_strength": 1.0,
|
| 1204 |
+
"kl/scaled_loss": 0.5429372787475586,
|
| 1205 |
+
"step": 350
|
| 1206 |
+
},
|
| 1207 |
+
{
|
| 1208 |
+
"epoch": 0.71,
|
| 1209 |
+
"learning_rate": 4.245562130177515e-05,
|
| 1210 |
+
"loss": 1.0536,
|
| 1211 |
+
"step": 355
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 0.71,
|
| 1215 |
+
"kl/loss": 0.5574665665626526,
|
| 1216 |
+
"kl/reg_strength": 1.0,
|
| 1217 |
+
"kl/scaled_loss": 0.5574665665626526,
|
| 1218 |
+
"step": 355
|
| 1219 |
+
},
|
| 1220 |
+
{
|
| 1221 |
+
"epoch": 0.72,
|
| 1222 |
+
"learning_rate": 4.2270710059171596e-05,
|
| 1223 |
+
"loss": 0.9456,
|
| 1224 |
+
"step": 360
|
| 1225 |
+
},
|
| 1226 |
+
{
|
| 1227 |
+
"epoch": 0.72,
|
| 1228 |
+
"eval_loss": 0.5382819771766663,
|
| 1229 |
+
"eval_runtime": 4.8554,
|
| 1230 |
+
"eval_samples_per_second": 102.978,
|
| 1231 |
+
"eval_steps_per_second": 25.744,
|
| 1232 |
+
"step": 360
|
| 1233 |
+
},
|
| 1234 |
+
{
|
| 1235 |
+
"epoch": 0.72,
|
| 1236 |
+
"kl/loss": 0.3959071636199951,
|
| 1237 |
+
"kl/reg_strength": 1.0,
|
| 1238 |
+
"kl/scaled_loss": 0.3959071636199951,
|
| 1239 |
+
"step": 360
|
| 1240 |
+
},
|
| 1241 |
+
{
|
| 1242 |
+
"epoch": 0.73,
|
| 1243 |
+
"learning_rate": 4.208579881656805e-05,
|
| 1244 |
+
"loss": 0.9666,
|
| 1245 |
+
"step": 365
|
| 1246 |
+
},
|
| 1247 |
+
{
|
| 1248 |
+
"epoch": 0.73,
|
| 1249 |
+
"kl/loss": 0.23654571175575256,
|
| 1250 |
+
"kl/reg_strength": 1.0,
|
| 1251 |
+
"kl/scaled_loss": 0.23654571175575256,
|
| 1252 |
+
"step": 365
|
| 1253 |
+
},
|
| 1254 |
+
{
|
| 1255 |
+
"epoch": 0.74,
|
| 1256 |
+
"learning_rate": 4.1900887573964496e-05,
|
| 1257 |
+
"loss": 1.0309,
|
| 1258 |
+
"step": 370
|
| 1259 |
+
},
|
| 1260 |
+
{
|
| 1261 |
+
"epoch": 0.74,
|
| 1262 |
+
"eval_loss": 0.5395053625106812,
|
| 1263 |
+
"eval_runtime": 4.8923,
|
| 1264 |
+
"eval_samples_per_second": 102.202,
|
| 1265 |
+
"eval_steps_per_second": 25.551,
|
| 1266 |
+
"step": 370
|
| 1267 |
+
},
|
| 1268 |
+
{
|
| 1269 |
+
"epoch": 0.74,
|
| 1270 |
+
"step": 370,
|
| 1271 |
+
"total_flos": 1127500686950400.0,
|
| 1272 |
+
"train_loss": 5.281058074332572,
|
| 1273 |
+
"train_runtime": 1534.4744,
|
| 1274 |
+
"train_samples_per_second": 7.826,
|
| 1275 |
+
"train_steps_per_second": 0.979
|
| 1276 |
+
}
|
| 1277 |
+
],
|
| 1278 |
+
"max_steps": 1503,
|
| 1279 |
+
"num_train_epochs": 3,
|
| 1280 |
+
"total_flos": 1127500686950400.0,
|
| 1281 |
+
"trial_name": null,
|
| 1282 |
+
"trial_params": null
|
| 1283 |
+
}
|
training_args.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 3695
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:227fe1adc3fa3c49c714785174a2296d7b2da1ba71d6520514f9e9a73248f0ab
|
| 3 |
size 3695
|