Upload task output 1280f28a-02cf-40d7-a301-7678a6e5aafd
Browse files- .gitattributes +1 -0
- added_tokens.json +6 -0
- config.json +30 -0
- generation_config.json +8 -0
- loss.txt +1 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- special_tokens_map.json +20 -0
- tokenizer.json +3 -0
- tokenizer_config.json +53 -0
- trainer_state.json +2727 -0
- training_args.bin +3 -0
- vocab.json +0 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
added_tokens.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"<|PAD_TOKEN|>": 151646,
|
| 3 |
+
"<|endoftext|>": 151643,
|
| 4 |
+
"<|im_end|>": 151645,
|
| 5 |
+
"<|im_start|>": 151644
|
| 6 |
+
}
|
config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen2ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_dropout": 0.0,
|
| 6 |
+
"bos_token_id": 151643,
|
| 7 |
+
"eos_token_id": 151643,
|
| 8 |
+
"hidden_act": "silu",
|
| 9 |
+
"hidden_size": 1536,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 8960,
|
| 12 |
+
"max_position_embeddings": 131072,
|
| 13 |
+
"max_window_layers": 28,
|
| 14 |
+
"model_type": "qwen2",
|
| 15 |
+
"num_attention_heads": 12,
|
| 16 |
+
"num_hidden_layers": 28,
|
| 17 |
+
"num_key_value_heads": 2,
|
| 18 |
+
"pad_token_id": 151646,
|
| 19 |
+
"rms_norm_eps": 1e-06,
|
| 20 |
+
"rope_scaling": null,
|
| 21 |
+
"rope_theta": 1000000.0,
|
| 22 |
+
"sliding_window": null,
|
| 23 |
+
"tie_word_embeddings": true,
|
| 24 |
+
"torch_dtype": "bfloat16",
|
| 25 |
+
"transformers_version": "4.51.3",
|
| 26 |
+
"unsloth_version": "2024.9",
|
| 27 |
+
"use_cache": false,
|
| 28 |
+
"use_sliding_window": false,
|
| 29 |
+
"vocab_size": 151936
|
| 30 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"eos_token_id": 151643,
|
| 4 |
+
"max_length": 131072,
|
| 5 |
+
"max_new_tokens": 2048,
|
| 6 |
+
"pad_token_id": 151646,
|
| 7 |
+
"transformers_version": "4.51.3"
|
| 8 |
+
}
|
loss.txt
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
1896,0.28488120436668396
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9596c70f607b3effd76ae3260f45213a2126e8481a9e5635b16179a053ae7009
|
| 3 |
+
size 3087467144
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|im_start|>",
|
| 4 |
+
"<|im_end|>"
|
| 5 |
+
],
|
| 6 |
+
"eos_token": {
|
| 7 |
+
"content": "<|endoftext|>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false
|
| 12 |
+
},
|
| 13 |
+
"pad_token": {
|
| 14 |
+
"content": "<|PAD_TOKEN|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false
|
| 19 |
+
}
|
| 20 |
+
}
|
tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9479047c22047670077878de944c696032b1a81049beeb6e99fbda7fb93e395a
|
| 3 |
+
size 11418456
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,53 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_prefix_space": false,
|
| 3 |
+
"added_tokens_decoder": {
|
| 4 |
+
"151643": {
|
| 5 |
+
"content": "<|endoftext|>",
|
| 6 |
+
"lstrip": false,
|
| 7 |
+
"normalized": false,
|
| 8 |
+
"rstrip": false,
|
| 9 |
+
"single_word": false,
|
| 10 |
+
"special": true
|
| 11 |
+
},
|
| 12 |
+
"151644": {
|
| 13 |
+
"content": "<|im_start|>",
|
| 14 |
+
"lstrip": false,
|
| 15 |
+
"normalized": false,
|
| 16 |
+
"rstrip": false,
|
| 17 |
+
"single_word": false,
|
| 18 |
+
"special": true
|
| 19 |
+
},
|
| 20 |
+
"151645": {
|
| 21 |
+
"content": "<|im_end|>",
|
| 22 |
+
"lstrip": false,
|
| 23 |
+
"normalized": false,
|
| 24 |
+
"rstrip": false,
|
| 25 |
+
"single_word": false,
|
| 26 |
+
"special": true
|
| 27 |
+
},
|
| 28 |
+
"151646": {
|
| 29 |
+
"content": "<|PAD_TOKEN|>",
|
| 30 |
+
"lstrip": false,
|
| 31 |
+
"normalized": false,
|
| 32 |
+
"rstrip": false,
|
| 33 |
+
"single_word": false,
|
| 34 |
+
"special": true
|
| 35 |
+
}
|
| 36 |
+
},
|
| 37 |
+
"additional_special_tokens": [
|
| 38 |
+
"<|im_start|>",
|
| 39 |
+
"<|im_end|>"
|
| 40 |
+
],
|
| 41 |
+
"bos_token": null,
|
| 42 |
+
"chat_template": "{% for message in messages %}{% if loop.first and messages[0]['role'] != 'system' %}{{ '<|im_start|>system\nYou are a helpful assistant<|im_end|>\n' }}{% endif %}{{'<|im_start|>' + message['role'] + '\n' + message['content'] + '<|im_end|>' + '\n'}}{% endfor %}{% if add_generation_prompt %}{{ '<|im_start|>assistant\n' }}{% endif %}",
|
| 43 |
+
"clean_up_tokenization_spaces": false,
|
| 44 |
+
"eos_token": "<|endoftext|>",
|
| 45 |
+
"errors": "replace",
|
| 46 |
+
"extra_special_tokens": {},
|
| 47 |
+
"model_max_length": 32768,
|
| 48 |
+
"pad_token": "<|PAD_TOKEN|>",
|
| 49 |
+
"padding_side": "left",
|
| 50 |
+
"split_special_tokens": false,
|
| 51 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 52 |
+
"unk_token": null
|
| 53 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2727 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 1.9978925184404637,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 1896,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.005268703898840885,
|
| 14 |
+
"grad_norm": 1.8046875,
|
| 15 |
+
"learning_rate": 9.142857142857142e-06,
|
| 16 |
+
"loss": 0.5469,
|
| 17 |
+
"step": 5
|
| 18 |
+
},
|
| 19 |
+
{
|
| 20 |
+
"epoch": 0.01053740779768177,
|
| 21 |
+
"grad_norm": 1.3203125,
|
| 22 |
+
"learning_rate": 2.0571428571428566e-05,
|
| 23 |
+
"loss": 0.493,
|
| 24 |
+
"step": 10
|
| 25 |
+
},
|
| 26 |
+
{
|
| 27 |
+
"epoch": 0.015806111696522657,
|
| 28 |
+
"grad_norm": 0.73828125,
|
| 29 |
+
"learning_rate": 3.2e-05,
|
| 30 |
+
"loss": 0.4305,
|
| 31 |
+
"step": 15
|
| 32 |
+
},
|
| 33 |
+
{
|
| 34 |
+
"epoch": 0.02107481559536354,
|
| 35 |
+
"grad_norm": 0.65625,
|
| 36 |
+
"learning_rate": 4.3428571428571424e-05,
|
| 37 |
+
"loss": 0.4097,
|
| 38 |
+
"step": 20
|
| 39 |
+
},
|
| 40 |
+
{
|
| 41 |
+
"epoch": 0.026343519494204427,
|
| 42 |
+
"grad_norm": 0.58984375,
|
| 43 |
+
"learning_rate": 5.485714285714285e-05,
|
| 44 |
+
"loss": 0.3943,
|
| 45 |
+
"step": 25
|
| 46 |
+
},
|
| 47 |
+
{
|
| 48 |
+
"epoch": 0.03161222339304531,
|
| 49 |
+
"grad_norm": 0.60546875,
|
| 50 |
+
"learning_rate": 6.628571428571428e-05,
|
| 51 |
+
"loss": 0.4013,
|
| 52 |
+
"step": 30
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"epoch": 0.0368809272918862,
|
| 56 |
+
"grad_norm": 0.640625,
|
| 57 |
+
"learning_rate": 7.771428571428571e-05,
|
| 58 |
+
"loss": 0.387,
|
| 59 |
+
"step": 35
|
| 60 |
+
},
|
| 61 |
+
{
|
| 62 |
+
"epoch": 0.04214963119072708,
|
| 63 |
+
"grad_norm": 0.6171875,
|
| 64 |
+
"learning_rate": 7.999970044289027e-05,
|
| 65 |
+
"loss": 0.3758,
|
| 66 |
+
"step": 40
|
| 67 |
+
},
|
| 68 |
+
{
|
| 69 |
+
"epoch": 0.04741833508956796,
|
| 70 |
+
"grad_norm": 0.609375,
|
| 71 |
+
"learning_rate": 7.999848350238486e-05,
|
| 72 |
+
"loss": 0.3781,
|
| 73 |
+
"step": 45
|
| 74 |
+
},
|
| 75 |
+
{
|
| 76 |
+
"epoch": 0.05268703898840885,
|
| 77 |
+
"grad_norm": 0.5546875,
|
| 78 |
+
"learning_rate": 7.999633049410842e-05,
|
| 79 |
+
"loss": 0.3766,
|
| 80 |
+
"step": 50
|
| 81 |
+
},
|
| 82 |
+
{
|
| 83 |
+
"epoch": 0.05795574288724974,
|
| 84 |
+
"grad_norm": 0.578125,
|
| 85 |
+
"learning_rate": 7.999324148524308e-05,
|
| 86 |
+
"loss": 0.3749,
|
| 87 |
+
"step": 55
|
| 88 |
+
},
|
| 89 |
+
{
|
| 90 |
+
"epoch": 0.06322444678609063,
|
| 91 |
+
"grad_norm": 0.5625,
|
| 92 |
+
"learning_rate": 7.998921657217774e-05,
|
| 93 |
+
"loss": 0.3719,
|
| 94 |
+
"step": 60
|
| 95 |
+
},
|
| 96 |
+
{
|
| 97 |
+
"epoch": 0.0684931506849315,
|
| 98 |
+
"grad_norm": 0.57421875,
|
| 99 |
+
"learning_rate": 7.998425588050514e-05,
|
| 100 |
+
"loss": 0.3694,
|
| 101 |
+
"step": 65
|
| 102 |
+
},
|
| 103 |
+
{
|
| 104 |
+
"epoch": 0.0737618545837724,
|
| 105 |
+
"grad_norm": 0.57421875,
|
| 106 |
+
"learning_rate": 7.99783595650179e-05,
|
| 107 |
+
"loss": 0.3679,
|
| 108 |
+
"step": 70
|
| 109 |
+
},
|
| 110 |
+
{
|
| 111 |
+
"epoch": 0.07903055848261328,
|
| 112 |
+
"grad_norm": 0.546875,
|
| 113 |
+
"learning_rate": 7.997152780970364e-05,
|
| 114 |
+
"loss": 0.3641,
|
| 115 |
+
"step": 75
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.08429926238145416,
|
| 119 |
+
"grad_norm": 0.625,
|
| 120 |
+
"learning_rate": 7.99637608277394e-05,
|
| 121 |
+
"loss": 0.3585,
|
| 122 |
+
"step": 80
|
| 123 |
+
},
|
| 124 |
+
{
|
| 125 |
+
"epoch": 0.08956796628029505,
|
| 126 |
+
"grad_norm": 0.55859375,
|
| 127 |
+
"learning_rate": 7.99550588614848e-05,
|
| 128 |
+
"loss": 0.3562,
|
| 129 |
+
"step": 85
|
| 130 |
+
},
|
| 131 |
+
{
|
| 132 |
+
"epoch": 0.09483667017913593,
|
| 133 |
+
"grad_norm": 0.50390625,
|
| 134 |
+
"learning_rate": 7.994542218247453e-05,
|
| 135 |
+
"loss": 0.3496,
|
| 136 |
+
"step": 90
|
| 137 |
+
},
|
| 138 |
+
{
|
| 139 |
+
"epoch": 0.10010537407797682,
|
| 140 |
+
"grad_norm": 0.56640625,
|
| 141 |
+
"learning_rate": 7.993485109140998e-05,
|
| 142 |
+
"loss": 0.3585,
|
| 143 |
+
"step": 95
|
| 144 |
+
},
|
| 145 |
+
{
|
| 146 |
+
"epoch": 0.1053740779768177,
|
| 147 |
+
"grad_norm": 0.55078125,
|
| 148 |
+
"learning_rate": 7.992334591814973e-05,
|
| 149 |
+
"loss": 0.3568,
|
| 150 |
+
"step": 100
|
| 151 |
+
},
|
| 152 |
+
{
|
| 153 |
+
"epoch": 0.11064278187565858,
|
| 154 |
+
"grad_norm": 0.53515625,
|
| 155 |
+
"learning_rate": 7.991090702169934e-05,
|
| 156 |
+
"loss": 0.349,
|
| 157 |
+
"step": 105
|
| 158 |
+
},
|
| 159 |
+
{
|
| 160 |
+
"epoch": 0.11591148577449947,
|
| 161 |
+
"grad_norm": 0.5234375,
|
| 162 |
+
"learning_rate": 7.989753479020009e-05,
|
| 163 |
+
"loss": 0.345,
|
| 164 |
+
"step": 110
|
| 165 |
+
},
|
| 166 |
+
{
|
| 167 |
+
"epoch": 0.12118018967334036,
|
| 168 |
+
"grad_norm": 0.5234375,
|
| 169 |
+
"learning_rate": 7.98832296409169e-05,
|
| 170 |
+
"loss": 0.3518,
|
| 171 |
+
"step": 115
|
| 172 |
+
},
|
| 173 |
+
{
|
| 174 |
+
"epoch": 0.12644889357218125,
|
| 175 |
+
"grad_norm": 0.5703125,
|
| 176 |
+
"learning_rate": 7.986799202022531e-05,
|
| 177 |
+
"loss": 0.357,
|
| 178 |
+
"step": 120
|
| 179 |
+
},
|
| 180 |
+
{
|
| 181 |
+
"epoch": 0.13171759747102213,
|
| 182 |
+
"grad_norm": 0.6328125,
|
| 183 |
+
"learning_rate": 7.985182240359757e-05,
|
| 184 |
+
"loss": 0.3563,
|
| 185 |
+
"step": 125
|
| 186 |
+
},
|
| 187 |
+
{
|
| 188 |
+
"epoch": 0.136986301369863,
|
| 189 |
+
"grad_norm": 0.515625,
|
| 190 |
+
"learning_rate": 7.983472129558769e-05,
|
| 191 |
+
"loss": 0.3465,
|
| 192 |
+
"step": 130
|
| 193 |
+
},
|
| 194 |
+
{
|
| 195 |
+
"epoch": 0.1422550052687039,
|
| 196 |
+
"grad_norm": 0.53515625,
|
| 197 |
+
"learning_rate": 7.98166892298159e-05,
|
| 198 |
+
"loss": 0.3397,
|
| 199 |
+
"step": 135
|
| 200 |
+
},
|
| 201 |
+
{
|
| 202 |
+
"epoch": 0.1475237091675448,
|
| 203 |
+
"grad_norm": 0.5390625,
|
| 204 |
+
"learning_rate": 7.979772676895182e-05,
|
| 205 |
+
"loss": 0.3541,
|
| 206 |
+
"step": 140
|
| 207 |
+
},
|
| 208 |
+
{
|
| 209 |
+
"epoch": 0.15279241306638566,
|
| 210 |
+
"grad_norm": 0.55859375,
|
| 211 |
+
"learning_rate": 7.977783450469697e-05,
|
| 212 |
+
"loss": 0.3431,
|
| 213 |
+
"step": 145
|
| 214 |
+
},
|
| 215 |
+
{
|
| 216 |
+
"epoch": 0.15806111696522657,
|
| 217 |
+
"grad_norm": 0.51171875,
|
| 218 |
+
"learning_rate": 7.97570130577663e-05,
|
| 219 |
+
"loss": 0.3507,
|
| 220 |
+
"step": 150
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.16332982086406744,
|
| 224 |
+
"grad_norm": 0.53515625,
|
| 225 |
+
"learning_rate": 7.973526307786885e-05,
|
| 226 |
+
"loss": 0.3416,
|
| 227 |
+
"step": 155
|
| 228 |
+
},
|
| 229 |
+
{
|
| 230 |
+
"epoch": 0.16859852476290832,
|
| 231 |
+
"grad_norm": 0.48828125,
|
| 232 |
+
"learning_rate": 7.971258524368743e-05,
|
| 233 |
+
"loss": 0.3459,
|
| 234 |
+
"step": 160
|
| 235 |
+
},
|
| 236 |
+
{
|
| 237 |
+
"epoch": 0.17386722866174922,
|
| 238 |
+
"grad_norm": 0.54296875,
|
| 239 |
+
"learning_rate": 7.968898026285744e-05,
|
| 240 |
+
"loss": 0.332,
|
| 241 |
+
"step": 165
|
| 242 |
+
},
|
| 243 |
+
{
|
| 244 |
+
"epoch": 0.1791359325605901,
|
| 245 |
+
"grad_norm": 0.52734375,
|
| 246 |
+
"learning_rate": 7.966444887194489e-05,
|
| 247 |
+
"loss": 0.3496,
|
| 248 |
+
"step": 170
|
| 249 |
+
},
|
| 250 |
+
{
|
| 251 |
+
"epoch": 0.18440463645943098,
|
| 252 |
+
"grad_norm": 0.515625,
|
| 253 |
+
"learning_rate": 7.963899183642324e-05,
|
| 254 |
+
"loss": 0.3409,
|
| 255 |
+
"step": 175
|
| 256 |
+
},
|
| 257 |
+
{
|
| 258 |
+
"epoch": 0.18967334035827185,
|
| 259 |
+
"grad_norm": 0.58984375,
|
| 260 |
+
"learning_rate": 7.961260995064969e-05,
|
| 261 |
+
"loss": 0.3473,
|
| 262 |
+
"step": 180
|
| 263 |
+
},
|
| 264 |
+
{
|
| 265 |
+
"epoch": 0.19494204425711276,
|
| 266 |
+
"grad_norm": 0.57421875,
|
| 267 |
+
"learning_rate": 7.958530403784029e-05,
|
| 268 |
+
"loss": 0.3389,
|
| 269 |
+
"step": 185
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"epoch": 0.20021074815595363,
|
| 273 |
+
"grad_norm": 0.51171875,
|
| 274 |
+
"learning_rate": 7.955707495004427e-05,
|
| 275 |
+
"loss": 0.3317,
|
| 276 |
+
"step": 190
|
| 277 |
+
},
|
| 278 |
+
{
|
| 279 |
+
"epoch": 0.2054794520547945,
|
| 280 |
+
"grad_norm": 0.52734375,
|
| 281 |
+
"learning_rate": 7.952792356811745e-05,
|
| 282 |
+
"loss": 0.3424,
|
| 283 |
+
"step": 195
|
| 284 |
+
},
|
| 285 |
+
{
|
| 286 |
+
"epoch": 0.2107481559536354,
|
| 287 |
+
"grad_norm": 0.5234375,
|
| 288 |
+
"learning_rate": 7.949785080169479e-05,
|
| 289 |
+
"loss": 0.3423,
|
| 290 |
+
"step": 200
|
| 291 |
+
},
|
| 292 |
+
{
|
| 293 |
+
"epoch": 0.2160168598524763,
|
| 294 |
+
"grad_norm": 0.490234375,
|
| 295 |
+
"learning_rate": 7.946685758916198e-05,
|
| 296 |
+
"loss": 0.338,
|
| 297 |
+
"step": 205
|
| 298 |
+
},
|
| 299 |
+
{
|
| 300 |
+
"epoch": 0.22128556375131717,
|
| 301 |
+
"grad_norm": 0.5234375,
|
| 302 |
+
"learning_rate": 7.943494489762617e-05,
|
| 303 |
+
"loss": 0.3256,
|
| 304 |
+
"step": 210
|
| 305 |
+
},
|
| 306 |
+
{
|
| 307 |
+
"epoch": 0.22655426765015807,
|
| 308 |
+
"grad_norm": 0.515625,
|
| 309 |
+
"learning_rate": 7.940211372288572e-05,
|
| 310 |
+
"loss": 0.3237,
|
| 311 |
+
"step": 215
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"epoch": 0.23182297154899895,
|
| 315 |
+
"grad_norm": 0.54296875,
|
| 316 |
+
"learning_rate": 7.936836508939928e-05,
|
| 317 |
+
"loss": 0.3334,
|
| 318 |
+
"step": 220
|
| 319 |
+
},
|
| 320 |
+
{
|
| 321 |
+
"epoch": 0.23709167544783982,
|
| 322 |
+
"grad_norm": 0.53125,
|
| 323 |
+
"learning_rate": 7.933370005025367e-05,
|
| 324 |
+
"loss": 0.3455,
|
| 325 |
+
"step": 225
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.24236037934668073,
|
| 329 |
+
"grad_norm": 0.53125,
|
| 330 |
+
"learning_rate": 7.92981196871311e-05,
|
| 331 |
+
"loss": 0.3303,
|
| 332 |
+
"step": 230
|
| 333 |
+
},
|
| 334 |
+
{
|
| 335 |
+
"epoch": 0.2476290832455216,
|
| 336 |
+
"grad_norm": 0.486328125,
|
| 337 |
+
"learning_rate": 7.926162511027539e-05,
|
| 338 |
+
"loss": 0.3389,
|
| 339 |
+
"step": 235
|
| 340 |
+
},
|
| 341 |
+
{
|
| 342 |
+
"epoch": 0.2528977871443625,
|
| 343 |
+
"grad_norm": 0.5,
|
| 344 |
+
"learning_rate": 7.922421745845734e-05,
|
| 345 |
+
"loss": 0.3347,
|
| 346 |
+
"step": 240
|
| 347 |
+
},
|
| 348 |
+
{
|
| 349 |
+
"epoch": 0.2581664910432034,
|
| 350 |
+
"grad_norm": 0.494140625,
|
| 351 |
+
"learning_rate": 7.918589789893922e-05,
|
| 352 |
+
"loss": 0.3215,
|
| 353 |
+
"step": 245
|
| 354 |
+
},
|
| 355 |
+
{
|
| 356 |
+
"epoch": 0.26343519494204426,
|
| 357 |
+
"grad_norm": 0.50390625,
|
| 358 |
+
"learning_rate": 7.914666762743831e-05,
|
| 359 |
+
"loss": 0.3234,
|
| 360 |
+
"step": 250
|
| 361 |
+
},
|
| 362 |
+
{
|
| 363 |
+
"epoch": 0.26870389884088514,
|
| 364 |
+
"grad_norm": 0.482421875,
|
| 365 |
+
"learning_rate": 7.910652786808953e-05,
|
| 366 |
+
"loss": 0.3211,
|
| 367 |
+
"step": 255
|
| 368 |
+
},
|
| 369 |
+
{
|
| 370 |
+
"epoch": 0.273972602739726,
|
| 371 |
+
"grad_norm": 0.4765625,
|
| 372 |
+
"learning_rate": 7.90654798734074e-05,
|
| 373 |
+
"loss": 0.3332,
|
| 374 |
+
"step": 260
|
| 375 |
+
},
|
| 376 |
+
{
|
| 377 |
+
"epoch": 0.2792413066385669,
|
| 378 |
+
"grad_norm": 0.486328125,
|
| 379 |
+
"learning_rate": 7.902352492424682e-05,
|
| 380 |
+
"loss": 0.3352,
|
| 381 |
+
"step": 265
|
| 382 |
+
},
|
| 383 |
+
{
|
| 384 |
+
"epoch": 0.2845100105374078,
|
| 385 |
+
"grad_norm": 0.484375,
|
| 386 |
+
"learning_rate": 7.898066432976318e-05,
|
| 387 |
+
"loss": 0.3403,
|
| 388 |
+
"step": 270
|
| 389 |
+
},
|
| 390 |
+
{
|
| 391 |
+
"epoch": 0.2897787144362487,
|
| 392 |
+
"grad_norm": 0.49609375,
|
| 393 |
+
"learning_rate": 7.893689942737141e-05,
|
| 394 |
+
"loss": 0.3351,
|
| 395 |
+
"step": 275
|
| 396 |
+
},
|
| 397 |
+
{
|
| 398 |
+
"epoch": 0.2950474183350896,
|
| 399 |
+
"grad_norm": 0.53515625,
|
| 400 |
+
"learning_rate": 7.88922315827044e-05,
|
| 401 |
+
"loss": 0.3254,
|
| 402 |
+
"step": 280
|
| 403 |
+
},
|
| 404 |
+
{
|
| 405 |
+
"epoch": 0.30031612223393045,
|
| 406 |
+
"grad_norm": 0.490234375,
|
| 407 |
+
"learning_rate": 7.884666218957029e-05,
|
| 408 |
+
"loss": 0.336,
|
| 409 |
+
"step": 285
|
| 410 |
+
},
|
| 411 |
+
{
|
| 412 |
+
"epoch": 0.3055848261327713,
|
| 413 |
+
"grad_norm": 0.47265625,
|
| 414 |
+
"learning_rate": 7.880019266990891e-05,
|
| 415 |
+
"loss": 0.318,
|
| 416 |
+
"step": 290
|
| 417 |
+
},
|
| 418 |
+
{
|
| 419 |
+
"epoch": 0.3108535300316122,
|
| 420 |
+
"grad_norm": 0.515625,
|
| 421 |
+
"learning_rate": 7.875282447374757e-05,
|
| 422 |
+
"loss": 0.3328,
|
| 423 |
+
"step": 295
|
| 424 |
+
},
|
| 425 |
+
{
|
| 426 |
+
"epoch": 0.31612223393045313,
|
| 427 |
+
"grad_norm": 0.5,
|
| 428 |
+
"learning_rate": 7.870455907915573e-05,
|
| 429 |
+
"loss": 0.3268,
|
| 430 |
+
"step": 300
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.321390937829294,
|
| 434 |
+
"grad_norm": 0.50390625,
|
| 435 |
+
"learning_rate": 7.865539799219885e-05,
|
| 436 |
+
"loss": 0.3188,
|
| 437 |
+
"step": 305
|
| 438 |
+
},
|
| 439 |
+
{
|
| 440 |
+
"epoch": 0.3266596417281349,
|
| 441 |
+
"grad_norm": 0.51171875,
|
| 442 |
+
"learning_rate": 7.860534274689147e-05,
|
| 443 |
+
"loss": 0.3286,
|
| 444 |
+
"step": 310
|
| 445 |
+
},
|
| 446 |
+
{
|
| 447 |
+
"epoch": 0.33192834562697576,
|
| 448 |
+
"grad_norm": 0.4609375,
|
| 449 |
+
"learning_rate": 7.855439490514922e-05,
|
| 450 |
+
"loss": 0.3259,
|
| 451 |
+
"step": 315
|
| 452 |
+
},
|
| 453 |
+
{
|
| 454 |
+
"epoch": 0.33719704952581664,
|
| 455 |
+
"grad_norm": 0.5,
|
| 456 |
+
"learning_rate": 7.850255605674026e-05,
|
| 457 |
+
"loss": 0.3213,
|
| 458 |
+
"step": 320
|
| 459 |
+
},
|
| 460 |
+
{
|
| 461 |
+
"epoch": 0.3424657534246575,
|
| 462 |
+
"grad_norm": 0.494140625,
|
| 463 |
+
"learning_rate": 7.844982781923554e-05,
|
| 464 |
+
"loss": 0.3194,
|
| 465 |
+
"step": 325
|
| 466 |
+
},
|
| 467 |
+
{
|
| 468 |
+
"epoch": 0.34773445732349845,
|
| 469 |
+
"grad_norm": 0.5390625,
|
| 470 |
+
"learning_rate": 7.839621183795833e-05,
|
| 471 |
+
"loss": 0.3258,
|
| 472 |
+
"step": 330
|
| 473 |
+
},
|
| 474 |
+
{
|
| 475 |
+
"epoch": 0.3530031612223393,
|
| 476 |
+
"grad_norm": 0.482421875,
|
| 477 |
+
"learning_rate": 7.834170978593296e-05,
|
| 478 |
+
"loss": 0.3232,
|
| 479 |
+
"step": 335
|
| 480 |
+
},
|
| 481 |
+
{
|
| 482 |
+
"epoch": 0.3582718651211802,
|
| 483 |
+
"grad_norm": 0.51953125,
|
| 484 |
+
"learning_rate": 7.828632336383253e-05,
|
| 485 |
+
"loss": 0.3232,
|
| 486 |
+
"step": 340
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"epoch": 0.3635405690200211,
|
| 490 |
+
"grad_norm": 0.4921875,
|
| 491 |
+
"learning_rate": 7.823005429992587e-05,
|
| 492 |
+
"loss": 0.3198,
|
| 493 |
+
"step": 345
|
| 494 |
+
},
|
| 495 |
+
{
|
| 496 |
+
"epoch": 0.36880927291886195,
|
| 497 |
+
"grad_norm": 0.4375,
|
| 498 |
+
"learning_rate": 7.81729043500237e-05,
|
| 499 |
+
"loss": 0.3222,
|
| 500 |
+
"step": 350
|
| 501 |
+
},
|
| 502 |
+
{
|
| 503 |
+
"epoch": 0.37407797681770283,
|
| 504 |
+
"grad_norm": 0.478515625,
|
| 505 |
+
"learning_rate": 7.811487529742366e-05,
|
| 506 |
+
"loss": 0.3269,
|
| 507 |
+
"step": 355
|
| 508 |
+
},
|
| 509 |
+
{
|
| 510 |
+
"epoch": 0.3793466807165437,
|
| 511 |
+
"grad_norm": 0.4765625,
|
| 512 |
+
"learning_rate": 7.805596895285485e-05,
|
| 513 |
+
"loss": 0.3221,
|
| 514 |
+
"step": 360
|
| 515 |
+
},
|
| 516 |
+
{
|
| 517 |
+
"epoch": 0.38461538461538464,
|
| 518 |
+
"grad_norm": 0.4921875,
|
| 519 |
+
"learning_rate": 7.799618715442116e-05,
|
| 520 |
+
"loss": 0.3178,
|
| 521 |
+
"step": 365
|
| 522 |
+
},
|
| 523 |
+
{
|
| 524 |
+
"epoch": 0.3898840885142255,
|
| 525 |
+
"grad_norm": 0.484375,
|
| 526 |
+
"learning_rate": 7.79355317675441e-05,
|
| 527 |
+
"loss": 0.3173,
|
| 528 |
+
"step": 370
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"epoch": 0.3951527924130664,
|
| 532 |
+
"grad_norm": 0.47265625,
|
| 533 |
+
"learning_rate": 7.78740046849044e-05,
|
| 534 |
+
"loss": 0.3206,
|
| 535 |
+
"step": 375
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.40042149631190727,
|
| 539 |
+
"grad_norm": 0.455078125,
|
| 540 |
+
"learning_rate": 7.781160782638307e-05,
|
| 541 |
+
"loss": 0.3141,
|
| 542 |
+
"step": 380
|
| 543 |
+
},
|
| 544 |
+
{
|
| 545 |
+
"epoch": 0.40569020021074814,
|
| 546 |
+
"grad_norm": 0.490234375,
|
| 547 |
+
"learning_rate": 7.774834313900154e-05,
|
| 548 |
+
"loss": 0.3107,
|
| 549 |
+
"step": 385
|
| 550 |
+
},
|
| 551 |
+
{
|
| 552 |
+
"epoch": 0.410958904109589,
|
| 553 |
+
"grad_norm": 0.474609375,
|
| 554 |
+
"learning_rate": 7.76842125968607e-05,
|
| 555 |
+
"loss": 0.3315,
|
| 556 |
+
"step": 390
|
| 557 |
+
},
|
| 558 |
+
{
|
| 559 |
+
"epoch": 0.41622760800842995,
|
| 560 |
+
"grad_norm": 0.474609375,
|
| 561 |
+
"learning_rate": 7.761921820107951e-05,
|
| 562 |
+
"loss": 0.315,
|
| 563 |
+
"step": 395
|
| 564 |
+
},
|
| 565 |
+
{
|
| 566 |
+
"epoch": 0.4214963119072708,
|
| 567 |
+
"grad_norm": 0.484375,
|
| 568 |
+
"learning_rate": 7.755336197973248e-05,
|
| 569 |
+
"loss": 0.311,
|
| 570 |
+
"step": 400
|
| 571 |
+
},
|
| 572 |
+
{
|
| 573 |
+
"epoch": 0.4267650158061117,
|
| 574 |
+
"grad_norm": 0.490234375,
|
| 575 |
+
"learning_rate": 7.748664598778633e-05,
|
| 576 |
+
"loss": 0.3257,
|
| 577 |
+
"step": 405
|
| 578 |
+
},
|
| 579 |
+
{
|
| 580 |
+
"epoch": 0.4320337197049526,
|
| 581 |
+
"grad_norm": 0.474609375,
|
| 582 |
+
"learning_rate": 7.7419072307036e-05,
|
| 583 |
+
"loss": 0.3256,
|
| 584 |
+
"step": 410
|
| 585 |
+
},
|
| 586 |
+
{
|
| 587 |
+
"epoch": 0.43730242360379346,
|
| 588 |
+
"grad_norm": 0.484375,
|
| 589 |
+
"learning_rate": 7.735064304603954e-05,
|
| 590 |
+
"loss": 0.3226,
|
| 591 |
+
"step": 415
|
| 592 |
+
},
|
| 593 |
+
{
|
| 594 |
+
"epoch": 0.44257112750263433,
|
| 595 |
+
"grad_norm": 0.458984375,
|
| 596 |
+
"learning_rate": 7.728136034005241e-05,
|
| 597 |
+
"loss": 0.3197,
|
| 598 |
+
"step": 420
|
| 599 |
+
},
|
| 600 |
+
{
|
| 601 |
+
"epoch": 0.44783983140147526,
|
| 602 |
+
"grad_norm": 0.458984375,
|
| 603 |
+
"learning_rate": 7.721122635096086e-05,
|
| 604 |
+
"loss": 0.3207,
|
| 605 |
+
"step": 425
|
| 606 |
+
},
|
| 607 |
+
{
|
| 608 |
+
"epoch": 0.45310853530031614,
|
| 609 |
+
"grad_norm": 0.462890625,
|
| 610 |
+
"learning_rate": 7.714024326721441e-05,
|
| 611 |
+
"loss": 0.3092,
|
| 612 |
+
"step": 430
|
| 613 |
+
},
|
| 614 |
+
{
|
| 615 |
+
"epoch": 0.458377239199157,
|
| 616 |
+
"grad_norm": 0.44140625,
|
| 617 |
+
"learning_rate": 7.706841330375755e-05,
|
| 618 |
+
"loss": 0.3104,
|
| 619 |
+
"step": 435
|
| 620 |
+
},
|
| 621 |
+
{
|
| 622 |
+
"epoch": 0.4636459430979979,
|
| 623 |
+
"grad_norm": 0.470703125,
|
| 624 |
+
"learning_rate": 7.699573870196074e-05,
|
| 625 |
+
"loss": 0.3036,
|
| 626 |
+
"step": 440
|
| 627 |
+
},
|
| 628 |
+
{
|
| 629 |
+
"epoch": 0.46891464699683877,
|
| 630 |
+
"grad_norm": 0.453125,
|
| 631 |
+
"learning_rate": 7.692222172955035e-05,
|
| 632 |
+
"loss": 0.3061,
|
| 633 |
+
"step": 445
|
| 634 |
+
},
|
| 635 |
+
{
|
| 636 |
+
"epoch": 0.47418335089567965,
|
| 637 |
+
"grad_norm": 0.474609375,
|
| 638 |
+
"learning_rate": 7.684786468053799e-05,
|
| 639 |
+
"loss": 0.3148,
|
| 640 |
+
"step": 450
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.4794520547945205,
|
| 644 |
+
"grad_norm": 0.466796875,
|
| 645 |
+
"learning_rate": 7.677266987514882e-05,
|
| 646 |
+
"loss": 0.3157,
|
| 647 |
+
"step": 455
|
| 648 |
+
},
|
| 649 |
+
{
|
| 650 |
+
"epoch": 0.48472075869336145,
|
| 651 |
+
"grad_norm": 0.4453125,
|
| 652 |
+
"learning_rate": 7.669663965974923e-05,
|
| 653 |
+
"loss": 0.3147,
|
| 654 |
+
"step": 460
|
| 655 |
+
},
|
| 656 |
+
{
|
| 657 |
+
"epoch": 0.48998946259220233,
|
| 658 |
+
"grad_norm": 0.44140625,
|
| 659 |
+
"learning_rate": 7.661977640677366e-05,
|
| 660 |
+
"loss": 0.3053,
|
| 661 |
+
"step": 465
|
| 662 |
+
},
|
| 663 |
+
{
|
| 664 |
+
"epoch": 0.4952581664910432,
|
| 665 |
+
"grad_norm": 0.431640625,
|
| 666 |
+
"learning_rate": 7.654208251465047e-05,
|
| 667 |
+
"loss": 0.3111,
|
| 668 |
+
"step": 470
|
| 669 |
+
},
|
| 670 |
+
{
|
| 671 |
+
"epoch": 0.5005268703898841,
|
| 672 |
+
"grad_norm": 0.4609375,
|
| 673 |
+
"learning_rate": 7.646356040772716e-05,
|
| 674 |
+
"loss": 0.3195,
|
| 675 |
+
"step": 475
|
| 676 |
+
},
|
| 677 |
+
{
|
| 678 |
+
"epoch": 0.505795574288725,
|
| 679 |
+
"grad_norm": 0.455078125,
|
| 680 |
+
"learning_rate": 7.638421253619466e-05,
|
| 681 |
+
"loss": 0.3026,
|
| 682 |
+
"step": 480
|
| 683 |
+
},
|
| 684 |
+
{
|
| 685 |
+
"epoch": 0.5110642781875658,
|
| 686 |
+
"grad_norm": 0.443359375,
|
| 687 |
+
"learning_rate": 7.630404137601104e-05,
|
| 688 |
+
"loss": 0.3004,
|
| 689 |
+
"step": 485
|
| 690 |
+
},
|
| 691 |
+
{
|
| 692 |
+
"epoch": 0.5163329820864068,
|
| 693 |
+
"grad_norm": 0.48828125,
|
| 694 |
+
"learning_rate": 7.622304942882402e-05,
|
| 695 |
+
"loss": 0.3181,
|
| 696 |
+
"step": 490
|
| 697 |
+
},
|
| 698 |
+
{
|
| 699 |
+
"epoch": 0.5216016859852476,
|
| 700 |
+
"grad_norm": 0.474609375,
|
| 701 |
+
"learning_rate": 7.61412392218931e-05,
|
| 702 |
+
"loss": 0.3061,
|
| 703 |
+
"step": 495
|
| 704 |
+
},
|
| 705 |
+
{
|
| 706 |
+
"epoch": 0.5268703898840885,
|
| 707 |
+
"grad_norm": 0.455078125,
|
| 708 |
+
"learning_rate": 7.605861330801056e-05,
|
| 709 |
+
"loss": 0.3244,
|
| 710 |
+
"step": 500
|
| 711 |
+
},
|
| 712 |
+
{
|
| 713 |
+
"epoch": 0.5268703898840885,
|
| 714 |
+
"eval_loss": 0.3330249488353729,
|
| 715 |
+
"eval_runtime": 2.6362,
|
| 716 |
+
"eval_samples_per_second": 17.829,
|
| 717 |
+
"eval_steps_per_second": 17.829,
|
| 718 |
+
"step": 500
|
| 719 |
+
},
|
| 720 |
+
{
|
| 721 |
+
"epoch": 0.5321390937829295,
|
| 722 |
+
"grad_norm": 0.466796875,
|
| 723 |
+
"learning_rate": 7.597517426542193e-05,
|
| 724 |
+
"loss": 0.3059,
|
| 725 |
+
"step": 505
|
| 726 |
+
},
|
| 727 |
+
{
|
| 728 |
+
"epoch": 0.5374077976817703,
|
| 729 |
+
"grad_norm": 0.458984375,
|
| 730 |
+
"learning_rate": 7.589092469774541e-05,
|
| 731 |
+
"loss": 0.307,
|
| 732 |
+
"step": 510
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"epoch": 0.5426765015806112,
|
| 736 |
+
"grad_norm": 0.435546875,
|
| 737 |
+
"learning_rate": 7.580586723389075e-05,
|
| 738 |
+
"loss": 0.3093,
|
| 739 |
+
"step": 515
|
| 740 |
+
},
|
| 741 |
+
{
|
| 742 |
+
"epoch": 0.547945205479452,
|
| 743 |
+
"grad_norm": 0.447265625,
|
| 744 |
+
"learning_rate": 7.572000452797713e-05,
|
| 745 |
+
"loss": 0.3079,
|
| 746 |
+
"step": 520
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 0.553213909378293,
|
| 750 |
+
"grad_norm": 0.453125,
|
| 751 |
+
"learning_rate": 7.563333925925036e-05,
|
| 752 |
+
"loss": 0.3077,
|
| 753 |
+
"step": 525
|
| 754 |
+
},
|
| 755 |
+
{
|
| 756 |
+
"epoch": 0.5584826132771338,
|
| 757 |
+
"grad_norm": 0.484375,
|
| 758 |
+
"learning_rate": 7.554587413199932e-05,
|
| 759 |
+
"loss": 0.3033,
|
| 760 |
+
"step": 530
|
| 761 |
+
},
|
| 762 |
+
{
|
| 763 |
+
"epoch": 0.5637513171759747,
|
| 764 |
+
"grad_norm": 0.423828125,
|
| 765 |
+
"learning_rate": 7.545761187547155e-05,
|
| 766 |
+
"loss": 0.3131,
|
| 767 |
+
"step": 535
|
| 768 |
+
},
|
| 769 |
+
{
|
| 770 |
+
"epoch": 0.5690200210748156,
|
| 771 |
+
"grad_norm": 0.421875,
|
| 772 |
+
"learning_rate": 7.536855524378804e-05,
|
| 773 |
+
"loss": 0.3053,
|
| 774 |
+
"step": 540
|
| 775 |
+
},
|
| 776 |
+
{
|
| 777 |
+
"epoch": 0.5742887249736565,
|
| 778 |
+
"grad_norm": 0.44921875,
|
| 779 |
+
"learning_rate": 7.527870701585735e-05,
|
| 780 |
+
"loss": 0.3038,
|
| 781 |
+
"step": 545
|
| 782 |
+
},
|
| 783 |
+
{
|
| 784 |
+
"epoch": 0.5795574288724974,
|
| 785 |
+
"grad_norm": 0.455078125,
|
| 786 |
+
"learning_rate": 7.518806999528887e-05,
|
| 787 |
+
"loss": 0.3086,
|
| 788 |
+
"step": 550
|
| 789 |
+
},
|
| 790 |
+
{
|
| 791 |
+
"epoch": 0.5848261327713382,
|
| 792 |
+
"grad_norm": 0.455078125,
|
| 793 |
+
"learning_rate": 7.50966470103054e-05,
|
| 794 |
+
"loss": 0.3003,
|
| 795 |
+
"step": 555
|
| 796 |
+
},
|
| 797 |
+
{
|
| 798 |
+
"epoch": 0.5900948366701791,
|
| 799 |
+
"grad_norm": 0.466796875,
|
| 800 |
+
"learning_rate": 7.500444091365479e-05,
|
| 801 |
+
"loss": 0.3078,
|
| 802 |
+
"step": 560
|
| 803 |
+
},
|
| 804 |
+
{
|
| 805 |
+
"epoch": 0.59536354056902,
|
| 806 |
+
"grad_norm": 0.44140625,
|
| 807 |
+
"learning_rate": 7.491145458252099e-05,
|
| 808 |
+
"loss": 0.3034,
|
| 809 |
+
"step": 565
|
| 810 |
+
},
|
| 811 |
+
{
|
| 812 |
+
"epoch": 0.6006322444678609,
|
| 813 |
+
"grad_norm": 0.47265625,
|
| 814 |
+
"learning_rate": 7.481769091843424e-05,
|
| 815 |
+
"loss": 0.3054,
|
| 816 |
+
"step": 570
|
| 817 |
+
},
|
| 818 |
+
{
|
| 819 |
+
"epoch": 0.6059009483667018,
|
| 820 |
+
"grad_norm": 0.44140625,
|
| 821 |
+
"learning_rate": 7.472315284718061e-05,
|
| 822 |
+
"loss": 0.2977,
|
| 823 |
+
"step": 575
|
| 824 |
+
},
|
| 825 |
+
{
|
| 826 |
+
"epoch": 0.6111696522655427,
|
| 827 |
+
"grad_norm": 0.427734375,
|
| 828 |
+
"learning_rate": 7.46278433187106e-05,
|
| 829 |
+
"loss": 0.2976,
|
| 830 |
+
"step": 580
|
| 831 |
+
},
|
| 832 |
+
{
|
| 833 |
+
"epoch": 0.6164383561643836,
|
| 834 |
+
"grad_norm": 0.4375,
|
| 835 |
+
"learning_rate": 7.453176530704713e-05,
|
| 836 |
+
"loss": 0.3064,
|
| 837 |
+
"step": 585
|
| 838 |
+
},
|
| 839 |
+
{
|
| 840 |
+
"epoch": 0.6217070600632244,
|
| 841 |
+
"grad_norm": 0.484375,
|
| 842 |
+
"learning_rate": 7.443492181019277e-05,
|
| 843 |
+
"loss": 0.2976,
|
| 844 |
+
"step": 590
|
| 845 |
+
},
|
| 846 |
+
{
|
| 847 |
+
"epoch": 0.6269757639620653,
|
| 848 |
+
"grad_norm": 0.455078125,
|
| 849 |
+
"learning_rate": 7.433731585003612e-05,
|
| 850 |
+
"loss": 0.2993,
|
| 851 |
+
"step": 595
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 0.6322444678609063,
|
| 855 |
+
"grad_norm": 0.4375,
|
| 856 |
+
"learning_rate": 7.423895047225762e-05,
|
| 857 |
+
"loss": 0.2986,
|
| 858 |
+
"step": 600
|
| 859 |
+
},
|
| 860 |
+
{
|
| 861 |
+
"epoch": 0.6375131717597471,
|
| 862 |
+
"grad_norm": 0.44140625,
|
| 863 |
+
"learning_rate": 7.413982874623443e-05,
|
| 864 |
+
"loss": 0.2996,
|
| 865 |
+
"step": 605
|
| 866 |
+
},
|
| 867 |
+
{
|
| 868 |
+
"epoch": 0.642781875658588,
|
| 869 |
+
"grad_norm": 0.462890625,
|
| 870 |
+
"learning_rate": 7.403995376494465e-05,
|
| 871 |
+
"loss": 0.3115,
|
| 872 |
+
"step": 610
|
| 873 |
+
},
|
| 874 |
+
{
|
| 875 |
+
"epoch": 0.6480505795574288,
|
| 876 |
+
"grad_norm": 0.455078125,
|
| 877 |
+
"learning_rate": 7.393932864487086e-05,
|
| 878 |
+
"loss": 0.3031,
|
| 879 |
+
"step": 615
|
| 880 |
+
},
|
| 881 |
+
{
|
| 882 |
+
"epoch": 0.6533192834562698,
|
| 883 |
+
"grad_norm": 0.462890625,
|
| 884 |
+
"learning_rate": 7.383795652590285e-05,
|
| 885 |
+
"loss": 0.3078,
|
| 886 |
+
"step": 620
|
| 887 |
+
},
|
| 888 |
+
{
|
| 889 |
+
"epoch": 0.6585879873551106,
|
| 890 |
+
"grad_norm": 0.455078125,
|
| 891 |
+
"learning_rate": 7.373584057123965e-05,
|
| 892 |
+
"loss": 0.2977,
|
| 893 |
+
"step": 625
|
| 894 |
+
},
|
| 895 |
+
{
|
| 896 |
+
"epoch": 0.6638566912539515,
|
| 897 |
+
"grad_norm": 0.41796875,
|
| 898 |
+
"learning_rate": 7.363298396729077e-05,
|
| 899 |
+
"loss": 0.2922,
|
| 900 |
+
"step": 630
|
| 901 |
+
},
|
| 902 |
+
{
|
| 903 |
+
"epoch": 0.6691253951527925,
|
| 904 |
+
"grad_norm": 0.4296875,
|
| 905 |
+
"learning_rate": 7.352938992357685e-05,
|
| 906 |
+
"loss": 0.2964,
|
| 907 |
+
"step": 635
|
| 908 |
+
},
|
| 909 |
+
{
|
| 910 |
+
"epoch": 0.6743940990516333,
|
| 911 |
+
"grad_norm": 0.44140625,
|
| 912 |
+
"learning_rate": 7.342506167262951e-05,
|
| 913 |
+
"loss": 0.3015,
|
| 914 |
+
"step": 640
|
| 915 |
+
},
|
| 916 |
+
{
|
| 917 |
+
"epoch": 0.6796628029504742,
|
| 918 |
+
"grad_norm": 0.443359375,
|
| 919 |
+
"learning_rate": 7.33200024698904e-05,
|
| 920 |
+
"loss": 0.3019,
|
| 921 |
+
"step": 645
|
| 922 |
+
},
|
| 923 |
+
{
|
| 924 |
+
"epoch": 0.684931506849315,
|
| 925 |
+
"grad_norm": 0.478515625,
|
| 926 |
+
"learning_rate": 7.32142155936097e-05,
|
| 927 |
+
"loss": 0.3052,
|
| 928 |
+
"step": 650
|
| 929 |
+
},
|
| 930 |
+
{
|
| 931 |
+
"epoch": 0.690200210748156,
|
| 932 |
+
"grad_norm": 0.439453125,
|
| 933 |
+
"learning_rate": 7.310770434474381e-05,
|
| 934 |
+
"loss": 0.2968,
|
| 935 |
+
"step": 655
|
| 936 |
+
},
|
| 937 |
+
{
|
| 938 |
+
"epoch": 0.6954689146469969,
|
| 939 |
+
"grad_norm": 0.439453125,
|
| 940 |
+
"learning_rate": 7.300047204685228e-05,
|
| 941 |
+
"loss": 0.3029,
|
| 942 |
+
"step": 660
|
| 943 |
+
},
|
| 944 |
+
{
|
| 945 |
+
"epoch": 0.7007376185458377,
|
| 946 |
+
"grad_norm": 0.421875,
|
| 947 |
+
"learning_rate": 7.28925220459942e-05,
|
| 948 |
+
"loss": 0.3056,
|
| 949 |
+
"step": 665
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"epoch": 0.7060063224446786,
|
| 953 |
+
"grad_norm": 0.451171875,
|
| 954 |
+
"learning_rate": 7.278385771062373e-05,
|
| 955 |
+
"loss": 0.3053,
|
| 956 |
+
"step": 670
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 0.7112750263435195,
|
| 960 |
+
"grad_norm": 0.44921875,
|
| 961 |
+
"learning_rate": 7.267448243148501e-05,
|
| 962 |
+
"loss": 0.2986,
|
| 963 |
+
"step": 675
|
| 964 |
+
},
|
| 965 |
+
{
|
| 966 |
+
"epoch": 0.7165437302423604,
|
| 967 |
+
"grad_norm": 0.462890625,
|
| 968 |
+
"learning_rate": 7.256439962150638e-05,
|
| 969 |
+
"loss": 0.2891,
|
| 970 |
+
"step": 680
|
| 971 |
+
},
|
| 972 |
+
{
|
| 973 |
+
"epoch": 0.7218124341412012,
|
| 974 |
+
"grad_norm": 0.416015625,
|
| 975 |
+
"learning_rate": 7.245361271569382e-05,
|
| 976 |
+
"loss": 0.2968,
|
| 977 |
+
"step": 685
|
| 978 |
+
},
|
| 979 |
+
{
|
| 980 |
+
"epoch": 0.7270811380400422,
|
| 981 |
+
"grad_norm": 0.455078125,
|
| 982 |
+
"learning_rate": 7.234212517102378e-05,
|
| 983 |
+
"loss": 0.3014,
|
| 984 |
+
"step": 690
|
| 985 |
+
},
|
| 986 |
+
{
|
| 987 |
+
"epoch": 0.7323498419388831,
|
| 988 |
+
"grad_norm": 0.48828125,
|
| 989 |
+
"learning_rate": 7.222994046633541e-05,
|
| 990 |
+
"loss": 0.3031,
|
| 991 |
+
"step": 695
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"epoch": 0.7376185458377239,
|
| 995 |
+
"grad_norm": 0.451171875,
|
| 996 |
+
"learning_rate": 7.211706210222186e-05,
|
| 997 |
+
"loss": 0.3031,
|
| 998 |
+
"step": 700
|
| 999 |
+
},
|
| 1000 |
+
{
|
| 1001 |
+
"epoch": 0.7428872497365648,
|
| 1002 |
+
"grad_norm": 0.421875,
|
| 1003 |
+
"learning_rate": 7.200349360092113e-05,
|
| 1004 |
+
"loss": 0.2889,
|
| 1005 |
+
"step": 705
|
| 1006 |
+
},
|
| 1007 |
+
{
|
| 1008 |
+
"epoch": 0.7481559536354057,
|
| 1009 |
+
"grad_norm": 0.427734375,
|
| 1010 |
+
"learning_rate": 7.188923850620616e-05,
|
| 1011 |
+
"loss": 0.2839,
|
| 1012 |
+
"step": 710
|
| 1013 |
+
},
|
| 1014 |
+
{
|
| 1015 |
+
"epoch": 0.7534246575342466,
|
| 1016 |
+
"grad_norm": 0.435546875,
|
| 1017 |
+
"learning_rate": 7.177430038327424e-05,
|
| 1018 |
+
"loss": 0.3007,
|
| 1019 |
+
"step": 715
|
| 1020 |
+
},
|
| 1021 |
+
{
|
| 1022 |
+
"epoch": 0.7586933614330874,
|
| 1023 |
+
"grad_norm": 0.431640625,
|
| 1024 |
+
"learning_rate": 7.165868281863572e-05,
|
| 1025 |
+
"loss": 0.2841,
|
| 1026 |
+
"step": 720
|
| 1027 |
+
},
|
| 1028 |
+
{
|
| 1029 |
+
"epoch": 0.7639620653319283,
|
| 1030 |
+
"grad_norm": 0.427734375,
|
| 1031 |
+
"learning_rate": 7.15423894200022e-05,
|
| 1032 |
+
"loss": 0.2979,
|
| 1033 |
+
"step": 725
|
| 1034 |
+
},
|
| 1035 |
+
{
|
| 1036 |
+
"epoch": 0.7692307692307693,
|
| 1037 |
+
"grad_norm": 0.435546875,
|
| 1038 |
+
"learning_rate": 7.142542381617388e-05,
|
| 1039 |
+
"loss": 0.2918,
|
| 1040 |
+
"step": 730
|
| 1041 |
+
},
|
| 1042 |
+
{
|
| 1043 |
+
"epoch": 0.7744994731296101,
|
| 1044 |
+
"grad_norm": 0.44140625,
|
| 1045 |
+
"learning_rate": 7.130778965692629e-05,
|
| 1046 |
+
"loss": 0.3029,
|
| 1047 |
+
"step": 735
|
| 1048 |
+
},
|
| 1049 |
+
{
|
| 1050 |
+
"epoch": 0.779768177028451,
|
| 1051 |
+
"grad_norm": 0.453125,
|
| 1052 |
+
"learning_rate": 7.118949061289649e-05,
|
| 1053 |
+
"loss": 0.3013,
|
| 1054 |
+
"step": 740
|
| 1055 |
+
},
|
| 1056 |
+
{
|
| 1057 |
+
"epoch": 0.7850368809272918,
|
| 1058 |
+
"grad_norm": 0.43359375,
|
| 1059 |
+
"learning_rate": 7.107053037546851e-05,
|
| 1060 |
+
"loss": 0.3047,
|
| 1061 |
+
"step": 745
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 0.7903055848261328,
|
| 1065 |
+
"grad_norm": 0.439453125,
|
| 1066 |
+
"learning_rate": 7.095091265665814e-05,
|
| 1067 |
+
"loss": 0.2979,
|
| 1068 |
+
"step": 750
|
| 1069 |
+
},
|
| 1070 |
+
{
|
| 1071 |
+
"epoch": 0.7955742887249737,
|
| 1072 |
+
"grad_norm": 0.466796875,
|
| 1073 |
+
"learning_rate": 7.083064118899708e-05,
|
| 1074 |
+
"loss": 0.293,
|
| 1075 |
+
"step": 755
|
| 1076 |
+
},
|
| 1077 |
+
{
|
| 1078 |
+
"epoch": 0.8008429926238145,
|
| 1079 |
+
"grad_norm": 0.431640625,
|
| 1080 |
+
"learning_rate": 7.070971972541654e-05,
|
| 1081 |
+
"loss": 0.2998,
|
| 1082 |
+
"step": 760
|
| 1083 |
+
},
|
| 1084 |
+
{
|
| 1085 |
+
"epoch": 0.8061116965226555,
|
| 1086 |
+
"grad_norm": 0.427734375,
|
| 1087 |
+
"learning_rate": 7.05881520391301e-05,
|
| 1088 |
+
"loss": 0.2864,
|
| 1089 |
+
"step": 765
|
| 1090 |
+
},
|
| 1091 |
+
{
|
| 1092 |
+
"epoch": 0.8113804004214963,
|
| 1093 |
+
"grad_norm": 0.42578125,
|
| 1094 |
+
"learning_rate": 7.046594192351595e-05,
|
| 1095 |
+
"loss": 0.2927,
|
| 1096 |
+
"step": 770
|
| 1097 |
+
},
|
| 1098 |
+
{
|
| 1099 |
+
"epoch": 0.8166491043203372,
|
| 1100 |
+
"grad_norm": 0.4453125,
|
| 1101 |
+
"learning_rate": 7.034309319199853e-05,
|
| 1102 |
+
"loss": 0.294,
|
| 1103 |
+
"step": 775
|
| 1104 |
+
},
|
| 1105 |
+
{
|
| 1106 |
+
"epoch": 0.821917808219178,
|
| 1107 |
+
"grad_norm": 0.4375,
|
| 1108 |
+
"learning_rate": 7.021960967792956e-05,
|
| 1109 |
+
"loss": 0.2912,
|
| 1110 |
+
"step": 780
|
| 1111 |
+
},
|
| 1112 |
+
{
|
| 1113 |
+
"epoch": 0.827186512118019,
|
| 1114 |
+
"grad_norm": 0.435546875,
|
| 1115 |
+
"learning_rate": 7.009549523446842e-05,
|
| 1116 |
+
"loss": 0.292,
|
| 1117 |
+
"step": 785
|
| 1118 |
+
},
|
| 1119 |
+
{
|
| 1120 |
+
"epoch": 0.8324552160168599,
|
| 1121 |
+
"grad_norm": 0.439453125,
|
| 1122 |
+
"learning_rate": 6.997075373446187e-05,
|
| 1123 |
+
"loss": 0.3041,
|
| 1124 |
+
"step": 790
|
| 1125 |
+
},
|
| 1126 |
+
{
|
| 1127 |
+
"epoch": 0.8377239199157007,
|
| 1128 |
+
"grad_norm": 0.435546875,
|
| 1129 |
+
"learning_rate": 6.98453890703233e-05,
|
| 1130 |
+
"loss": 0.2996,
|
| 1131 |
+
"step": 795
|
| 1132 |
+
},
|
| 1133 |
+
{
|
| 1134 |
+
"epoch": 0.8429926238145417,
|
| 1135 |
+
"grad_norm": 0.443359375,
|
| 1136 |
+
"learning_rate": 6.971940515391118e-05,
|
| 1137 |
+
"loss": 0.2953,
|
| 1138 |
+
"step": 800
|
| 1139 |
+
},
|
| 1140 |
+
{
|
| 1141 |
+
"epoch": 0.8482613277133825,
|
| 1142 |
+
"grad_norm": 0.43359375,
|
| 1143 |
+
"learning_rate": 6.9592805916407e-05,
|
| 1144 |
+
"loss": 0.2954,
|
| 1145 |
+
"step": 805
|
| 1146 |
+
},
|
| 1147 |
+
{
|
| 1148 |
+
"epoch": 0.8535300316122234,
|
| 1149 |
+
"grad_norm": 0.447265625,
|
| 1150 |
+
"learning_rate": 6.946559530819265e-05,
|
| 1151 |
+
"loss": 0.2937,
|
| 1152 |
+
"step": 810
|
| 1153 |
+
},
|
| 1154 |
+
{
|
| 1155 |
+
"epoch": 0.8587987355110642,
|
| 1156 |
+
"grad_norm": 0.44921875,
|
| 1157 |
+
"learning_rate": 6.933777729872716e-05,
|
| 1158 |
+
"loss": 0.3035,
|
| 1159 |
+
"step": 815
|
| 1160 |
+
},
|
| 1161 |
+
{
|
| 1162 |
+
"epoch": 0.8640674394099052,
|
| 1163 |
+
"grad_norm": 0.435546875,
|
| 1164 |
+
"learning_rate": 6.920935587642278e-05,
|
| 1165 |
+
"loss": 0.2913,
|
| 1166 |
+
"step": 820
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 0.8693361433087461,
|
| 1170 |
+
"grad_norm": 0.435546875,
|
| 1171 |
+
"learning_rate": 6.908033504852054e-05,
|
| 1172 |
+
"loss": 0.2961,
|
| 1173 |
+
"step": 825
|
| 1174 |
+
},
|
| 1175 |
+
{
|
| 1176 |
+
"epoch": 0.8746048472075869,
|
| 1177 |
+
"grad_norm": 0.43359375,
|
| 1178 |
+
"learning_rate": 6.895071884096526e-05,
|
| 1179 |
+
"loss": 0.2846,
|
| 1180 |
+
"step": 830
|
| 1181 |
+
},
|
| 1182 |
+
{
|
| 1183 |
+
"epoch": 0.8798735511064278,
|
| 1184 |
+
"grad_norm": 0.423828125,
|
| 1185 |
+
"learning_rate": 6.882051129827989e-05,
|
| 1186 |
+
"loss": 0.288,
|
| 1187 |
+
"step": 835
|
| 1188 |
+
},
|
| 1189 |
+
{
|
| 1190 |
+
"epoch": 0.8851422550052687,
|
| 1191 |
+
"grad_norm": 0.4453125,
|
| 1192 |
+
"learning_rate": 6.868971648343925e-05,
|
| 1193 |
+
"loss": 0.2927,
|
| 1194 |
+
"step": 840
|
| 1195 |
+
},
|
| 1196 |
+
{
|
| 1197 |
+
"epoch": 0.8904109589041096,
|
| 1198 |
+
"grad_norm": 0.423828125,
|
| 1199 |
+
"learning_rate": 6.855833847774337e-05,
|
| 1200 |
+
"loss": 0.286,
|
| 1201 |
+
"step": 845
|
| 1202 |
+
},
|
| 1203 |
+
{
|
| 1204 |
+
"epoch": 0.8956796628029505,
|
| 1205 |
+
"grad_norm": 0.447265625,
|
| 1206 |
+
"learning_rate": 6.842638138069003e-05,
|
| 1207 |
+
"loss": 0.2863,
|
| 1208 |
+
"step": 850
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"epoch": 0.9009483667017913,
|
| 1212 |
+
"grad_norm": 0.42578125,
|
| 1213 |
+
"learning_rate": 6.82938493098469e-05,
|
| 1214 |
+
"loss": 0.2901,
|
| 1215 |
+
"step": 855
|
| 1216 |
+
},
|
| 1217 |
+
{
|
| 1218 |
+
"epoch": 0.9062170706006323,
|
| 1219 |
+
"grad_norm": 0.439453125,
|
| 1220 |
+
"learning_rate": 6.816074640072305e-05,
|
| 1221 |
+
"loss": 0.2884,
|
| 1222 |
+
"step": 860
|
| 1223 |
+
},
|
| 1224 |
+
{
|
| 1225 |
+
"epoch": 0.9114857744994731,
|
| 1226 |
+
"grad_norm": 0.4375,
|
| 1227 |
+
"learning_rate": 6.802707680663987e-05,
|
| 1228 |
+
"loss": 0.2903,
|
| 1229 |
+
"step": 865
|
| 1230 |
+
},
|
| 1231 |
+
{
|
| 1232 |
+
"epoch": 0.916754478398314,
|
| 1233 |
+
"grad_norm": 0.43359375,
|
| 1234 |
+
"learning_rate": 6.789284469860146e-05,
|
| 1235 |
+
"loss": 0.2945,
|
| 1236 |
+
"step": 870
|
| 1237 |
+
},
|
| 1238 |
+
{
|
| 1239 |
+
"epoch": 0.9220231822971549,
|
| 1240 |
+
"grad_norm": 0.44140625,
|
| 1241 |
+
"learning_rate": 6.775805426516464e-05,
|
| 1242 |
+
"loss": 0.2864,
|
| 1243 |
+
"step": 875
|
| 1244 |
+
},
|
| 1245 |
+
{
|
| 1246 |
+
"epoch": 0.9272918861959958,
|
| 1247 |
+
"grad_norm": 0.439453125,
|
| 1248 |
+
"learning_rate": 6.762270971230799e-05,
|
| 1249 |
+
"loss": 0.287,
|
| 1250 |
+
"step": 880
|
| 1251 |
+
},
|
| 1252 |
+
{
|
| 1253 |
+
"epoch": 0.9325605900948367,
|
| 1254 |
+
"grad_norm": 0.44140625,
|
| 1255 |
+
"learning_rate": 6.748681526330078e-05,
|
| 1256 |
+
"loss": 0.2936,
|
| 1257 |
+
"step": 885
|
| 1258 |
+
},
|
| 1259 |
+
{
|
| 1260 |
+
"epoch": 0.9378292939936775,
|
| 1261 |
+
"grad_norm": 0.4375,
|
| 1262 |
+
"learning_rate": 6.73503751585712e-05,
|
| 1263 |
+
"loss": 0.2907,
|
| 1264 |
+
"step": 890
|
| 1265 |
+
},
|
| 1266 |
+
{
|
| 1267 |
+
"epoch": 0.9430979978925185,
|
| 1268 |
+
"grad_norm": 0.43359375,
|
| 1269 |
+
"learning_rate": 6.721339365557394e-05,
|
| 1270 |
+
"loss": 0.2864,
|
| 1271 |
+
"step": 895
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 0.9483667017913593,
|
| 1275 |
+
"grad_norm": 0.419921875,
|
| 1276 |
+
"learning_rate": 6.707587502865739e-05,
|
| 1277 |
+
"loss": 0.2948,
|
| 1278 |
+
"step": 900
|
| 1279 |
+
},
|
| 1280 |
+
{
|
| 1281 |
+
"epoch": 0.9536354056902002,
|
| 1282 |
+
"grad_norm": 0.4375,
|
| 1283 |
+
"learning_rate": 6.693782356893032e-05,
|
| 1284 |
+
"loss": 0.2962,
|
| 1285 |
+
"step": 905
|
| 1286 |
+
},
|
| 1287 |
+
{
|
| 1288 |
+
"epoch": 0.958904109589041,
|
| 1289 |
+
"grad_norm": 0.439453125,
|
| 1290 |
+
"learning_rate": 6.679924358412785e-05,
|
| 1291 |
+
"loss": 0.2898,
|
| 1292 |
+
"step": 910
|
| 1293 |
+
},
|
| 1294 |
+
{
|
| 1295 |
+
"epoch": 0.964172813487882,
|
| 1296 |
+
"grad_norm": 0.451171875,
|
| 1297 |
+
"learning_rate": 6.666013939847719e-05,
|
| 1298 |
+
"loss": 0.2833,
|
| 1299 |
+
"step": 915
|
| 1300 |
+
},
|
| 1301 |
+
{
|
| 1302 |
+
"epoch": 0.9694415173867229,
|
| 1303 |
+
"grad_norm": 0.44140625,
|
| 1304 |
+
"learning_rate": 6.652051535256257e-05,
|
| 1305 |
+
"loss": 0.2995,
|
| 1306 |
+
"step": 920
|
| 1307 |
+
},
|
| 1308 |
+
{
|
| 1309 |
+
"epoch": 0.9747102212855637,
|
| 1310 |
+
"grad_norm": 0.44140625,
|
| 1311 |
+
"learning_rate": 6.638037580318988e-05,
|
| 1312 |
+
"loss": 0.2863,
|
| 1313 |
+
"step": 925
|
| 1314 |
+
},
|
| 1315 |
+
{
|
| 1316 |
+
"epoch": 0.9799789251844047,
|
| 1317 |
+
"grad_norm": 0.478515625,
|
| 1318 |
+
"learning_rate": 6.623972512325068e-05,
|
| 1319 |
+
"loss": 0.294,
|
| 1320 |
+
"step": 930
|
| 1321 |
+
},
|
| 1322 |
+
{
|
| 1323 |
+
"epoch": 0.9852476290832455,
|
| 1324 |
+
"grad_norm": 0.41015625,
|
| 1325 |
+
"learning_rate": 6.609856770158579e-05,
|
| 1326 |
+
"loss": 0.2806,
|
| 1327 |
+
"step": 935
|
| 1328 |
+
},
|
| 1329 |
+
{
|
| 1330 |
+
"epoch": 0.9905163329820864,
|
| 1331 |
+
"grad_norm": 0.421875,
|
| 1332 |
+
"learning_rate": 6.595690794284828e-05,
|
| 1333 |
+
"loss": 0.2901,
|
| 1334 |
+
"step": 940
|
| 1335 |
+
},
|
| 1336 |
+
{
|
| 1337 |
+
"epoch": 0.9957850368809273,
|
| 1338 |
+
"grad_norm": 0.400390625,
|
| 1339 |
+
"learning_rate": 6.581475026736611e-05,
|
| 1340 |
+
"loss": 0.2799,
|
| 1341 |
+
"step": 945
|
| 1342 |
+
},
|
| 1343 |
+
{
|
| 1344 |
+
"epoch": 0.9989462592202318,
|
| 1345 |
+
"eval_loss": 0.3042255640029907,
|
| 1346 |
+
"eval_runtime": 2.5158,
|
| 1347 |
+
"eval_samples_per_second": 18.682,
|
| 1348 |
+
"eval_steps_per_second": 18.682,
|
| 1349 |
+
"step": 948
|
| 1350 |
+
},
|
| 1351 |
+
{
|
| 1352 |
+
"epoch": 1.0010537407797682,
|
| 1353 |
+
"grad_norm": 0.59765625,
|
| 1354 |
+
"learning_rate": 6.56720991110041e-05,
|
| 1355 |
+
"loss": 0.2765,
|
| 1356 |
+
"step": 950
|
| 1357 |
+
},
|
| 1358 |
+
{
|
| 1359 |
+
"epoch": 1.006322444678609,
|
| 1360 |
+
"grad_norm": 0.51171875,
|
| 1361 |
+
"learning_rate": 6.552895892502563e-05,
|
| 1362 |
+
"loss": 0.2285,
|
| 1363 |
+
"step": 955
|
| 1364 |
+
},
|
| 1365 |
+
{
|
| 1366 |
+
"epoch": 1.01159114857745,
|
| 1367 |
+
"grad_norm": 0.416015625,
|
| 1368 |
+
"learning_rate": 6.538533417595359e-05,
|
| 1369 |
+
"loss": 0.2271,
|
| 1370 |
+
"step": 960
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"epoch": 1.0168598524762908,
|
| 1374 |
+
"grad_norm": 0.451171875,
|
| 1375 |
+
"learning_rate": 6.52412293454312e-05,
|
| 1376 |
+
"loss": 0.2292,
|
| 1377 |
+
"step": 965
|
| 1378 |
+
},
|
| 1379 |
+
{
|
| 1380 |
+
"epoch": 1.0221285563751317,
|
| 1381 |
+
"grad_norm": 0.44140625,
|
| 1382 |
+
"learning_rate": 6.5096648930082e-05,
|
| 1383 |
+
"loss": 0.2196,
|
| 1384 |
+
"step": 970
|
| 1385 |
+
},
|
| 1386 |
+
{
|
| 1387 |
+
"epoch": 1.0273972602739727,
|
| 1388 |
+
"grad_norm": 0.439453125,
|
| 1389 |
+
"learning_rate": 6.495159744136959e-05,
|
| 1390 |
+
"loss": 0.2249,
|
| 1391 |
+
"step": 975
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 1.0326659641728135,
|
| 1395 |
+
"grad_norm": 0.431640625,
|
| 1396 |
+
"learning_rate": 6.480607940545692e-05,
|
| 1397 |
+
"loss": 0.2212,
|
| 1398 |
+
"step": 980
|
| 1399 |
+
},
|
| 1400 |
+
{
|
| 1401 |
+
"epoch": 1.0379346680716544,
|
| 1402 |
+
"grad_norm": 0.44921875,
|
| 1403 |
+
"learning_rate": 6.466009936306498e-05,
|
| 1404 |
+
"loss": 0.2272,
|
| 1405 |
+
"step": 985
|
| 1406 |
+
},
|
| 1407 |
+
{
|
| 1408 |
+
"epoch": 1.0432033719704952,
|
| 1409 |
+
"grad_norm": 0.408203125,
|
| 1410 |
+
"learning_rate": 6.45136618693311e-05,
|
| 1411 |
+
"loss": 0.2253,
|
| 1412 |
+
"step": 990
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"epoch": 1.0484720758693362,
|
| 1416 |
+
"grad_norm": 0.447265625,
|
| 1417 |
+
"learning_rate": 6.436677149366688e-05,
|
| 1418 |
+
"loss": 0.229,
|
| 1419 |
+
"step": 995
|
| 1420 |
+
},
|
| 1421 |
+
{
|
| 1422 |
+
"epoch": 1.053740779768177,
|
| 1423 |
+
"grad_norm": 0.427734375,
|
| 1424 |
+
"learning_rate": 6.42194328196156e-05,
|
| 1425 |
+
"loss": 0.2252,
|
| 1426 |
+
"step": 1000
|
| 1427 |
+
},
|
| 1428 |
+
{
|
| 1429 |
+
"epoch": 1.053740779768177,
|
| 1430 |
+
"eval_loss": 0.30919456481933594,
|
| 1431 |
+
"eval_runtime": 2.4891,
|
| 1432 |
+
"eval_samples_per_second": 18.882,
|
| 1433 |
+
"eval_steps_per_second": 18.882,
|
| 1434 |
+
"step": 1000
|
| 1435 |
+
},
|
| 1436 |
+
{
|
| 1437 |
+
"epoch": 1.0590094836670179,
|
| 1438 |
+
"grad_norm": 0.494140625,
|
| 1439 |
+
"learning_rate": 6.407165044470911e-05,
|
| 1440 |
+
"loss": 0.2241,
|
| 1441 |
+
"step": 1005
|
| 1442 |
+
},
|
| 1443 |
+
{
|
| 1444 |
+
"epoch": 1.064278187565859,
|
| 1445 |
+
"grad_norm": 0.4375,
|
| 1446 |
+
"learning_rate": 6.392342898032445e-05,
|
| 1447 |
+
"loss": 0.2296,
|
| 1448 |
+
"step": 1010
|
| 1449 |
+
},
|
| 1450 |
+
{
|
| 1451 |
+
"epoch": 1.0695468914646997,
|
| 1452 |
+
"grad_norm": 0.478515625,
|
| 1453 |
+
"learning_rate": 6.377477305153997e-05,
|
| 1454 |
+
"loss": 0.2257,
|
| 1455 |
+
"step": 1015
|
| 1456 |
+
},
|
| 1457 |
+
{
|
| 1458 |
+
"epoch": 1.0748155953635405,
|
| 1459 |
+
"grad_norm": 0.478515625,
|
| 1460 |
+
"learning_rate": 6.362568729699093e-05,
|
| 1461 |
+
"loss": 0.2269,
|
| 1462 |
+
"step": 1020
|
| 1463 |
+
},
|
| 1464 |
+
{
|
| 1465 |
+
"epoch": 1.0800842992623814,
|
| 1466 |
+
"grad_norm": 0.48828125,
|
| 1467 |
+
"learning_rate": 6.347617636872484e-05,
|
| 1468 |
+
"loss": 0.2314,
|
| 1469 |
+
"step": 1025
|
| 1470 |
+
},
|
| 1471 |
+
{
|
| 1472 |
+
"epoch": 1.0853530031612224,
|
| 1473 |
+
"grad_norm": 0.439453125,
|
| 1474 |
+
"learning_rate": 6.332624493205623e-05,
|
| 1475 |
+
"loss": 0.2261,
|
| 1476 |
+
"step": 1030
|
| 1477 |
+
},
|
| 1478 |
+
{
|
| 1479 |
+
"epoch": 1.0906217070600632,
|
| 1480 |
+
"grad_norm": 0.45703125,
|
| 1481 |
+
"learning_rate": 6.317589766542111e-05,
|
| 1482 |
+
"loss": 0.2239,
|
| 1483 |
+
"step": 1035
|
| 1484 |
+
},
|
| 1485 |
+
{
|
| 1486 |
+
"epoch": 1.095890410958904,
|
| 1487 |
+
"grad_norm": 0.431640625,
|
| 1488 |
+
"learning_rate": 6.3025139260231e-05,
|
| 1489 |
+
"loss": 0.2234,
|
| 1490 |
+
"step": 1040
|
| 1491 |
+
},
|
| 1492 |
+
{
|
| 1493 |
+
"epoch": 1.101159114857745,
|
| 1494 |
+
"grad_norm": 0.4453125,
|
| 1495 |
+
"learning_rate": 6.287397442072653e-05,
|
| 1496 |
+
"loss": 0.2342,
|
| 1497 |
+
"step": 1045
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 1.106427818756586,
|
| 1501 |
+
"grad_norm": 0.443359375,
|
| 1502 |
+
"learning_rate": 6.272240786383057e-05,
|
| 1503 |
+
"loss": 0.2307,
|
| 1504 |
+
"step": 1050
|
| 1505 |
+
},
|
| 1506 |
+
{
|
| 1507 |
+
"epoch": 1.1116965226554267,
|
| 1508 |
+
"grad_norm": 0.443359375,
|
| 1509 |
+
"learning_rate": 6.257044431900121e-05,
|
| 1510 |
+
"loss": 0.2376,
|
| 1511 |
+
"step": 1055
|
| 1512 |
+
},
|
| 1513 |
+
{
|
| 1514 |
+
"epoch": 1.1169652265542676,
|
| 1515 |
+
"grad_norm": 0.4375,
|
| 1516 |
+
"learning_rate": 6.241808852808403e-05,
|
| 1517 |
+
"loss": 0.2319,
|
| 1518 |
+
"step": 1060
|
| 1519 |
+
},
|
| 1520 |
+
{
|
| 1521 |
+
"epoch": 1.1222339304531086,
|
| 1522 |
+
"grad_norm": 0.439453125,
|
| 1523 |
+
"learning_rate": 6.226534524516418e-05,
|
| 1524 |
+
"loss": 0.2285,
|
| 1525 |
+
"step": 1065
|
| 1526 |
+
},
|
| 1527 |
+
{
|
| 1528 |
+
"epoch": 1.1275026343519494,
|
| 1529 |
+
"grad_norm": 0.427734375,
|
| 1530 |
+
"learning_rate": 6.21122192364181e-05,
|
| 1531 |
+
"loss": 0.2254,
|
| 1532 |
+
"step": 1070
|
| 1533 |
+
},
|
| 1534 |
+
{
|
| 1535 |
+
"epoch": 1.1327713382507902,
|
| 1536 |
+
"grad_norm": 0.4453125,
|
| 1537 |
+
"learning_rate": 6.195871527996467e-05,
|
| 1538 |
+
"loss": 0.2325,
|
| 1539 |
+
"step": 1075
|
| 1540 |
+
},
|
| 1541 |
+
{
|
| 1542 |
+
"epoch": 1.1380400421496313,
|
| 1543 |
+
"grad_norm": 0.421875,
|
| 1544 |
+
"learning_rate": 6.180483816571628e-05,
|
| 1545 |
+
"loss": 0.2333,
|
| 1546 |
+
"step": 1080
|
| 1547 |
+
},
|
| 1548 |
+
{
|
| 1549 |
+
"epoch": 1.143308746048472,
|
| 1550 |
+
"grad_norm": 0.41796875,
|
| 1551 |
+
"learning_rate": 6.165059269522921e-05,
|
| 1552 |
+
"loss": 0.2162,
|
| 1553 |
+
"step": 1085
|
| 1554 |
+
},
|
| 1555 |
+
{
|
| 1556 |
+
"epoch": 1.148577449947313,
|
| 1557 |
+
"grad_norm": 0.44140625,
|
| 1558 |
+
"learning_rate": 6.149598368155386e-05,
|
| 1559 |
+
"loss": 0.2269,
|
| 1560 |
+
"step": 1090
|
| 1561 |
+
},
|
| 1562 |
+
{
|
| 1563 |
+
"epoch": 1.1538461538461537,
|
| 1564 |
+
"grad_norm": 0.419921875,
|
| 1565 |
+
"learning_rate": 6.13410159490846e-05,
|
| 1566 |
+
"loss": 0.2304,
|
| 1567 |
+
"step": 1095
|
| 1568 |
+
},
|
| 1569 |
+
{
|
| 1570 |
+
"epoch": 1.1591148577449948,
|
| 1571 |
+
"grad_norm": 0.48046875,
|
| 1572 |
+
"learning_rate": 6.118569433340927e-05,
|
| 1573 |
+
"loss": 0.2246,
|
| 1574 |
+
"step": 1100
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"epoch": 1.1643835616438356,
|
| 1578 |
+
"grad_norm": 0.427734375,
|
| 1579 |
+
"learning_rate": 6.103002368115805e-05,
|
| 1580 |
+
"loss": 0.219,
|
| 1581 |
+
"step": 1105
|
| 1582 |
+
},
|
| 1583 |
+
{
|
| 1584 |
+
"epoch": 1.1696522655426764,
|
| 1585 |
+
"grad_norm": 0.43359375,
|
| 1586 |
+
"learning_rate": 6.0874008849852566e-05,
|
| 1587 |
+
"loss": 0.2299,
|
| 1588 |
+
"step": 1110
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 1.1749209694415175,
|
| 1592 |
+
"grad_norm": 0.4296875,
|
| 1593 |
+
"learning_rate": 6.071765470775406e-05,
|
| 1594 |
+
"loss": 0.2222,
|
| 1595 |
+
"step": 1115
|
| 1596 |
+
},
|
| 1597 |
+
{
|
| 1598 |
+
"epoch": 1.1801896733403583,
|
| 1599 |
+
"grad_norm": 0.423828125,
|
| 1600 |
+
"learning_rate": 6.056096613371163e-05,
|
| 1601 |
+
"loss": 0.224,
|
| 1602 |
+
"step": 1120
|
| 1603 |
+
},
|
| 1604 |
+
{
|
| 1605 |
+
"epoch": 1.1854583772391991,
|
| 1606 |
+
"grad_norm": 0.458984375,
|
| 1607 |
+
"learning_rate": 6.040394801700989e-05,
|
| 1608 |
+
"loss": 0.2249,
|
| 1609 |
+
"step": 1125
|
| 1610 |
+
},
|
| 1611 |
+
{
|
| 1612 |
+
"epoch": 1.1907270811380402,
|
| 1613 |
+
"grad_norm": 0.4375,
|
| 1614 |
+
"learning_rate": 6.024660525721645e-05,
|
| 1615 |
+
"loss": 0.2188,
|
| 1616 |
+
"step": 1130
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"epoch": 1.195995785036881,
|
| 1620 |
+
"grad_norm": 0.423828125,
|
| 1621 |
+
"learning_rate": 6.008894276402905e-05,
|
| 1622 |
+
"loss": 0.2212,
|
| 1623 |
+
"step": 1135
|
| 1624 |
+
},
|
| 1625 |
+
{
|
| 1626 |
+
"epoch": 1.2012644889357218,
|
| 1627 |
+
"grad_norm": 0.4375,
|
| 1628 |
+
"learning_rate": 5.993096545712233e-05,
|
| 1629 |
+
"loss": 0.2234,
|
| 1630 |
+
"step": 1140
|
| 1631 |
+
},
|
| 1632 |
+
{
|
| 1633 |
+
"epoch": 1.2065331928345626,
|
| 1634 |
+
"grad_norm": 0.42578125,
|
| 1635 |
+
"learning_rate": 5.977267826599435e-05,
|
| 1636 |
+
"loss": 0.2186,
|
| 1637 |
+
"step": 1145
|
| 1638 |
+
},
|
| 1639 |
+
{
|
| 1640 |
+
"epoch": 1.2118018967334037,
|
| 1641 |
+
"grad_norm": 0.4453125,
|
| 1642 |
+
"learning_rate": 5.9614086129812696e-05,
|
| 1643 |
+
"loss": 0.2269,
|
| 1644 |
+
"step": 1150
|
| 1645 |
+
},
|
| 1646 |
+
{
|
| 1647 |
+
"epoch": 1.2170706006322445,
|
| 1648 |
+
"grad_norm": 0.4609375,
|
| 1649 |
+
"learning_rate": 5.945519399726045e-05,
|
| 1650 |
+
"loss": 0.2257,
|
| 1651 |
+
"step": 1155
|
| 1652 |
+
},
|
| 1653 |
+
{
|
| 1654 |
+
"epoch": 1.2223393045310853,
|
| 1655 |
+
"grad_norm": 0.443359375,
|
| 1656 |
+
"learning_rate": 5.929600682638171e-05,
|
| 1657 |
+
"loss": 0.2274,
|
| 1658 |
+
"step": 1160
|
| 1659 |
+
},
|
| 1660 |
+
{
|
| 1661 |
+
"epoch": 1.2276080084299261,
|
| 1662 |
+
"grad_norm": 0.4296875,
|
| 1663 |
+
"learning_rate": 5.913652958442693e-05,
|
| 1664 |
+
"loss": 0.2169,
|
| 1665 |
+
"step": 1165
|
| 1666 |
+
},
|
| 1667 |
+
{
|
| 1668 |
+
"epoch": 1.2328767123287672,
|
| 1669 |
+
"grad_norm": 0.44921875,
|
| 1670 |
+
"learning_rate": 5.8976767247697856e-05,
|
| 1671 |
+
"loss": 0.2267,
|
| 1672 |
+
"step": 1170
|
| 1673 |
+
},
|
| 1674 |
+
{
|
| 1675 |
+
"epoch": 1.238145416227608,
|
| 1676 |
+
"grad_norm": 0.4296875,
|
| 1677 |
+
"learning_rate": 5.88167248013923e-05,
|
| 1678 |
+
"loss": 0.2205,
|
| 1679 |
+
"step": 1175
|
| 1680 |
+
},
|
| 1681 |
+
{
|
| 1682 |
+
"epoch": 1.2434141201264488,
|
| 1683 |
+
"grad_norm": 0.453125,
|
| 1684 |
+
"learning_rate": 5.865640723944859e-05,
|
| 1685 |
+
"loss": 0.2256,
|
| 1686 |
+
"step": 1180
|
| 1687 |
+
},
|
| 1688 |
+
{
|
| 1689 |
+
"epoch": 1.2486828240252899,
|
| 1690 |
+
"grad_norm": 0.44140625,
|
| 1691 |
+
"learning_rate": 5.849581956438969e-05,
|
| 1692 |
+
"loss": 0.2294,
|
| 1693 |
+
"step": 1185
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 1.2539515279241307,
|
| 1697 |
+
"grad_norm": 0.470703125,
|
| 1698 |
+
"learning_rate": 5.8334966787167135e-05,
|
| 1699 |
+
"loss": 0.2247,
|
| 1700 |
+
"step": 1190
|
| 1701 |
+
},
|
| 1702 |
+
{
|
| 1703 |
+
"epoch": 1.2592202318229715,
|
| 1704 |
+
"grad_norm": 0.419921875,
|
| 1705 |
+
"learning_rate": 5.8173853927004676e-05,
|
| 1706 |
+
"loss": 0.2187,
|
| 1707 |
+
"step": 1195
|
| 1708 |
+
},
|
| 1709 |
+
{
|
| 1710 |
+
"epoch": 1.2644889357218125,
|
| 1711 |
+
"grad_norm": 0.4453125,
|
| 1712 |
+
"learning_rate": 5.801248601124164e-05,
|
| 1713 |
+
"loss": 0.2276,
|
| 1714 |
+
"step": 1200
|
| 1715 |
+
},
|
| 1716 |
+
{
|
| 1717 |
+
"epoch": 1.2697576396206534,
|
| 1718 |
+
"grad_norm": 0.42578125,
|
| 1719 |
+
"learning_rate": 5.7850868075176056e-05,
|
| 1720 |
+
"loss": 0.235,
|
| 1721 |
+
"step": 1205
|
| 1722 |
+
},
|
| 1723 |
+
{
|
| 1724 |
+
"epoch": 1.2750263435194942,
|
| 1725 |
+
"grad_norm": 0.416015625,
|
| 1726 |
+
"learning_rate": 5.7689005161907586e-05,
|
| 1727 |
+
"loss": 0.2252,
|
| 1728 |
+
"step": 1210
|
| 1729 |
+
},
|
| 1730 |
+
{
|
| 1731 |
+
"epoch": 1.2802950474183352,
|
| 1732 |
+
"grad_norm": 0.4296875,
|
| 1733 |
+
"learning_rate": 5.752690232218005e-05,
|
| 1734 |
+
"loss": 0.2244,
|
| 1735 |
+
"step": 1215
|
| 1736 |
+
},
|
| 1737 |
+
{
|
| 1738 |
+
"epoch": 1.285563751317176,
|
| 1739 |
+
"grad_norm": 0.44921875,
|
| 1740 |
+
"learning_rate": 5.7364564614223974e-05,
|
| 1741 |
+
"loss": 0.2224,
|
| 1742 |
+
"step": 1220
|
| 1743 |
+
},
|
| 1744 |
+
{
|
| 1745 |
+
"epoch": 1.2908324552160169,
|
| 1746 |
+
"grad_norm": 0.44140625,
|
| 1747 |
+
"learning_rate": 5.720199710359862e-05,
|
| 1748 |
+
"loss": 0.2329,
|
| 1749 |
+
"step": 1225
|
| 1750 |
+
},
|
| 1751 |
+
{
|
| 1752 |
+
"epoch": 1.2961011591148577,
|
| 1753 |
+
"grad_norm": 0.419921875,
|
| 1754 |
+
"learning_rate": 5.703920486303399e-05,
|
| 1755 |
+
"loss": 0.2193,
|
| 1756 |
+
"step": 1230
|
| 1757 |
+
},
|
| 1758 |
+
{
|
| 1759 |
+
"epoch": 1.3013698630136985,
|
| 1760 |
+
"grad_norm": 0.44140625,
|
| 1761 |
+
"learning_rate": 5.6876192972272516e-05,
|
| 1762 |
+
"loss": 0.2306,
|
| 1763 |
+
"step": 1235
|
| 1764 |
+
},
|
| 1765 |
+
{
|
| 1766 |
+
"epoch": 1.3066385669125395,
|
| 1767 |
+
"grad_norm": 0.443359375,
|
| 1768 |
+
"learning_rate": 5.6712966517910595e-05,
|
| 1769 |
+
"loss": 0.2199,
|
| 1770 |
+
"step": 1240
|
| 1771 |
+
},
|
| 1772 |
+
{
|
| 1773 |
+
"epoch": 1.3119072708113804,
|
| 1774 |
+
"grad_norm": 0.42578125,
|
| 1775 |
+
"learning_rate": 5.654953059323978e-05,
|
| 1776 |
+
"loss": 0.2217,
|
| 1777 |
+
"step": 1245
|
| 1778 |
+
},
|
| 1779 |
+
{
|
| 1780 |
+
"epoch": 1.3171759747102212,
|
| 1781 |
+
"grad_norm": 0.42578125,
|
| 1782 |
+
"learning_rate": 5.638589029808793e-05,
|
| 1783 |
+
"loss": 0.2288,
|
| 1784 |
+
"step": 1250
|
| 1785 |
+
},
|
| 1786 |
+
{
|
| 1787 |
+
"epoch": 1.3224446786090622,
|
| 1788 |
+
"grad_norm": 0.45703125,
|
| 1789 |
+
"learning_rate": 5.6222050738660043e-05,
|
| 1790 |
+
"loss": 0.2265,
|
| 1791 |
+
"step": 1255
|
| 1792 |
+
},
|
| 1793 |
+
{
|
| 1794 |
+
"epoch": 1.327713382507903,
|
| 1795 |
+
"grad_norm": 0.431640625,
|
| 1796 |
+
"learning_rate": 5.605801702737892e-05,
|
| 1797 |
+
"loss": 0.2341,
|
| 1798 |
+
"step": 1260
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 1.3329820864067439,
|
| 1802 |
+
"grad_norm": 0.453125,
|
| 1803 |
+
"learning_rate": 5.5893794282725646e-05,
|
| 1804 |
+
"loss": 0.23,
|
| 1805 |
+
"step": 1265
|
| 1806 |
+
},
|
| 1807 |
+
{
|
| 1808 |
+
"epoch": 1.338250790305585,
|
| 1809 |
+
"grad_norm": 0.4453125,
|
| 1810 |
+
"learning_rate": 5.5729387629079884e-05,
|
| 1811 |
+
"loss": 0.2156,
|
| 1812 |
+
"step": 1270
|
| 1813 |
+
},
|
| 1814 |
+
{
|
| 1815 |
+
"epoch": 1.3435194942044257,
|
| 1816 |
+
"grad_norm": 0.4609375,
|
| 1817 |
+
"learning_rate": 5.556480219655995e-05,
|
| 1818 |
+
"loss": 0.2279,
|
| 1819 |
+
"step": 1275
|
| 1820 |
+
},
|
| 1821 |
+
{
|
| 1822 |
+
"epoch": 1.3487881981032666,
|
| 1823 |
+
"grad_norm": 0.4296875,
|
| 1824 |
+
"learning_rate": 5.540004312086276e-05,
|
| 1825 |
+
"loss": 0.2238,
|
| 1826 |
+
"step": 1280
|
| 1827 |
+
},
|
| 1828 |
+
{
|
| 1829 |
+
"epoch": 1.3540569020021076,
|
| 1830 |
+
"grad_norm": 0.474609375,
|
| 1831 |
+
"learning_rate": 5.523511554310354e-05,
|
| 1832 |
+
"loss": 0.234,
|
| 1833 |
+
"step": 1285
|
| 1834 |
+
},
|
| 1835 |
+
{
|
| 1836 |
+
"epoch": 1.3593256059009484,
|
| 1837 |
+
"grad_norm": 0.43359375,
|
| 1838 |
+
"learning_rate": 5.507002460965547e-05,
|
| 1839 |
+
"loss": 0.2243,
|
| 1840 |
+
"step": 1290
|
| 1841 |
+
},
|
| 1842 |
+
{
|
| 1843 |
+
"epoch": 1.3645943097997892,
|
| 1844 |
+
"grad_norm": 0.43359375,
|
| 1845 |
+
"learning_rate": 5.490477547198899e-05,
|
| 1846 |
+
"loss": 0.2217,
|
| 1847 |
+
"step": 1295
|
| 1848 |
+
},
|
| 1849 |
+
{
|
| 1850 |
+
"epoch": 1.36986301369863,
|
| 1851 |
+
"grad_norm": 0.455078125,
|
| 1852 |
+
"learning_rate": 5.47393732865112e-05,
|
| 1853 |
+
"loss": 0.2297,
|
| 1854 |
+
"step": 1300
|
| 1855 |
+
},
|
| 1856 |
+
{
|
| 1857 |
+
"epoch": 1.375131717597471,
|
| 1858 |
+
"grad_norm": 0.421875,
|
| 1859 |
+
"learning_rate": 5.457382321440477e-05,
|
| 1860 |
+
"loss": 0.2282,
|
| 1861 |
+
"step": 1305
|
| 1862 |
+
},
|
| 1863 |
+
{
|
| 1864 |
+
"epoch": 1.380400421496312,
|
| 1865 |
+
"grad_norm": 0.44921875,
|
| 1866 |
+
"learning_rate": 5.4408130421467115e-05,
|
| 1867 |
+
"loss": 0.2275,
|
| 1868 |
+
"step": 1310
|
| 1869 |
+
},
|
| 1870 |
+
{
|
| 1871 |
+
"epoch": 1.3856691253951527,
|
| 1872 |
+
"grad_norm": 0.43359375,
|
| 1873 |
+
"learning_rate": 5.424230007794903e-05,
|
| 1874 |
+
"loss": 0.2246,
|
| 1875 |
+
"step": 1315
|
| 1876 |
+
},
|
| 1877 |
+
{
|
| 1878 |
+
"epoch": 1.3909378292939936,
|
| 1879 |
+
"grad_norm": 0.486328125,
|
| 1880 |
+
"learning_rate": 5.40763373583934e-05,
|
| 1881 |
+
"loss": 0.2187,
|
| 1882 |
+
"step": 1320
|
| 1883 |
+
},
|
| 1884 |
+
{
|
| 1885 |
+
"epoch": 1.3962065331928346,
|
| 1886 |
+
"grad_norm": 0.45703125,
|
| 1887 |
+
"learning_rate": 5.391024744147379e-05,
|
| 1888 |
+
"loss": 0.2226,
|
| 1889 |
+
"step": 1325
|
| 1890 |
+
},
|
| 1891 |
+
{
|
| 1892 |
+
"epoch": 1.4014752370916754,
|
| 1893 |
+
"grad_norm": 0.455078125,
|
| 1894 |
+
"learning_rate": 5.374403550983279e-05,
|
| 1895 |
+
"loss": 0.2213,
|
| 1896 |
+
"step": 1330
|
| 1897 |
+
},
|
| 1898 |
+
{
|
| 1899 |
+
"epoch": 1.4067439409905163,
|
| 1900 |
+
"grad_norm": 0.455078125,
|
| 1901 |
+
"learning_rate": 5.357770674992032e-05,
|
| 1902 |
+
"loss": 0.226,
|
| 1903 |
+
"step": 1335
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 1.4120126448893573,
|
| 1907 |
+
"grad_norm": 0.42578125,
|
| 1908 |
+
"learning_rate": 5.341126635183178e-05,
|
| 1909 |
+
"loss": 0.2262,
|
| 1910 |
+
"step": 1340
|
| 1911 |
+
},
|
| 1912 |
+
{
|
| 1913 |
+
"epoch": 1.4172813487881981,
|
| 1914 |
+
"grad_norm": 0.44921875,
|
| 1915 |
+
"learning_rate": 5.324471950914613e-05,
|
| 1916 |
+
"loss": 0.2298,
|
| 1917 |
+
"step": 1345
|
| 1918 |
+
},
|
| 1919 |
+
{
|
| 1920 |
+
"epoch": 1.422550052687039,
|
| 1921 |
+
"grad_norm": 0.4453125,
|
| 1922 |
+
"learning_rate": 5.30780714187638e-05,
|
| 1923 |
+
"loss": 0.2197,
|
| 1924 |
+
"step": 1350
|
| 1925 |
+
},
|
| 1926 |
+
{
|
| 1927 |
+
"epoch": 1.42781875658588,
|
| 1928 |
+
"grad_norm": 0.421875,
|
| 1929 |
+
"learning_rate": 5.291132728074453e-05,
|
| 1930 |
+
"loss": 0.2231,
|
| 1931 |
+
"step": 1355
|
| 1932 |
+
},
|
| 1933 |
+
{
|
| 1934 |
+
"epoch": 1.4330874604847208,
|
| 1935 |
+
"grad_norm": 0.44140625,
|
| 1936 |
+
"learning_rate": 5.2744492298145136e-05,
|
| 1937 |
+
"loss": 0.2251,
|
| 1938 |
+
"step": 1360
|
| 1939 |
+
},
|
| 1940 |
+
{
|
| 1941 |
+
"epoch": 1.4383561643835616,
|
| 1942 |
+
"grad_norm": 0.453125,
|
| 1943 |
+
"learning_rate": 5.25775716768571e-05,
|
| 1944 |
+
"loss": 0.2252,
|
| 1945 |
+
"step": 1365
|
| 1946 |
+
},
|
| 1947 |
+
{
|
| 1948 |
+
"epoch": 1.4436248682824027,
|
| 1949 |
+
"grad_norm": 0.42578125,
|
| 1950 |
+
"learning_rate": 5.24105706254442e-05,
|
| 1951 |
+
"loss": 0.2238,
|
| 1952 |
+
"step": 1370
|
| 1953 |
+
},
|
| 1954 |
+
{
|
| 1955 |
+
"epoch": 1.4488935721812435,
|
| 1956 |
+
"grad_norm": 0.4140625,
|
| 1957 |
+
"learning_rate": 5.224349435497989e-05,
|
| 1958 |
+
"loss": 0.2259,
|
| 1959 |
+
"step": 1375
|
| 1960 |
+
},
|
| 1961 |
+
{
|
| 1962 |
+
"epoch": 1.4541622760800843,
|
| 1963 |
+
"grad_norm": 0.44140625,
|
| 1964 |
+
"learning_rate": 5.207634807888481e-05,
|
| 1965 |
+
"loss": 0.2172,
|
| 1966 |
+
"step": 1380
|
| 1967 |
+
},
|
| 1968 |
+
{
|
| 1969 |
+
"epoch": 1.4594309799789251,
|
| 1970 |
+
"grad_norm": 0.43359375,
|
| 1971 |
+
"learning_rate": 5.190913701276398e-05,
|
| 1972 |
+
"loss": 0.2133,
|
| 1973 |
+
"step": 1385
|
| 1974 |
+
},
|
| 1975 |
+
{
|
| 1976 |
+
"epoch": 1.464699683877766,
|
| 1977 |
+
"grad_norm": 0.43359375,
|
| 1978 |
+
"learning_rate": 5.1741866374244174e-05,
|
| 1979 |
+
"loss": 0.2195,
|
| 1980 |
+
"step": 1390
|
| 1981 |
+
},
|
| 1982 |
+
{
|
| 1983 |
+
"epoch": 1.469968387776607,
|
| 1984 |
+
"grad_norm": 0.4296875,
|
| 1985 |
+
"learning_rate": 5.157454138281102e-05,
|
| 1986 |
+
"loss": 0.2223,
|
| 1987 |
+
"step": 1395
|
| 1988 |
+
},
|
| 1989 |
+
{
|
| 1990 |
+
"epoch": 1.4752370916754478,
|
| 1991 |
+
"grad_norm": 0.431640625,
|
| 1992 |
+
"learning_rate": 5.140716725964618e-05,
|
| 1993 |
+
"loss": 0.2163,
|
| 1994 |
+
"step": 1400
|
| 1995 |
+
},
|
| 1996 |
+
{
|
| 1997 |
+
"epoch": 1.4805057955742886,
|
| 1998 |
+
"grad_norm": 0.4140625,
|
| 1999 |
+
"learning_rate": 5.1239749227464393e-05,
|
| 2000 |
+
"loss": 0.2264,
|
| 2001 |
+
"step": 1405
|
| 2002 |
+
},
|
| 2003 |
+
{
|
| 2004 |
+
"epoch": 1.4857744994731297,
|
| 2005 |
+
"grad_norm": 0.416015625,
|
| 2006 |
+
"learning_rate": 5.107229251035056e-05,
|
| 2007 |
+
"loss": 0.2168,
|
| 2008 |
+
"step": 1410
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 1.4910432033719705,
|
| 2012 |
+
"grad_norm": 0.4140625,
|
| 2013 |
+
"learning_rate": 5.090480233359667e-05,
|
| 2014 |
+
"loss": 0.2221,
|
| 2015 |
+
"step": 1415
|
| 2016 |
+
},
|
| 2017 |
+
{
|
| 2018 |
+
"epoch": 1.4963119072708113,
|
| 2019 |
+
"grad_norm": 0.435546875,
|
| 2020 |
+
"learning_rate": 5.07372839235388e-05,
|
| 2021 |
+
"loss": 0.2257,
|
| 2022 |
+
"step": 1420
|
| 2023 |
+
},
|
| 2024 |
+
{
|
| 2025 |
+
"epoch": 1.5015806111696524,
|
| 2026 |
+
"grad_norm": 0.443359375,
|
| 2027 |
+
"learning_rate": 5.056974250739401e-05,
|
| 2028 |
+
"loss": 0.2241,
|
| 2029 |
+
"step": 1425
|
| 2030 |
+
},
|
| 2031 |
+
{
|
| 2032 |
+
"epoch": 1.5068493150684932,
|
| 2033 |
+
"grad_norm": 0.451171875,
|
| 2034 |
+
"learning_rate": 5.0402183313097235e-05,
|
| 2035 |
+
"loss": 0.2253,
|
| 2036 |
+
"step": 1430
|
| 2037 |
+
},
|
| 2038 |
+
{
|
| 2039 |
+
"epoch": 1.512118018967334,
|
| 2040 |
+
"grad_norm": 0.4375,
|
| 2041 |
+
"learning_rate": 5.023461156913818e-05,
|
| 2042 |
+
"loss": 0.2195,
|
| 2043 |
+
"step": 1435
|
| 2044 |
+
},
|
| 2045 |
+
{
|
| 2046 |
+
"epoch": 1.517386722866175,
|
| 2047 |
+
"grad_norm": 0.43359375,
|
| 2048 |
+
"learning_rate": 5.0067032504398086e-05,
|
| 2049 |
+
"loss": 0.2172,
|
| 2050 |
+
"step": 1440
|
| 2051 |
+
},
|
| 2052 |
+
{
|
| 2053 |
+
"epoch": 1.5226554267650156,
|
| 2054 |
+
"grad_norm": 0.423828125,
|
| 2055 |
+
"learning_rate": 4.98994513479867e-05,
|
| 2056 |
+
"loss": 0.2243,
|
| 2057 |
+
"step": 1445
|
| 2058 |
+
},
|
| 2059 |
+
{
|
| 2060 |
+
"epoch": 1.5279241306638567,
|
| 2061 |
+
"grad_norm": 0.423828125,
|
| 2062 |
+
"learning_rate": 4.9731873329079e-05,
|
| 2063 |
+
"loss": 0.224,
|
| 2064 |
+
"step": 1450
|
| 2065 |
+
},
|
| 2066 |
+
{
|
| 2067 |
+
"epoch": 1.5331928345626977,
|
| 2068 |
+
"grad_norm": 0.404296875,
|
| 2069 |
+
"learning_rate": 4.9564303676752075e-05,
|
| 2070 |
+
"loss": 0.2187,
|
| 2071 |
+
"step": 1455
|
| 2072 |
+
},
|
| 2073 |
+
{
|
| 2074 |
+
"epoch": 1.5384615384615383,
|
| 2075 |
+
"grad_norm": 0.44140625,
|
| 2076 |
+
"learning_rate": 4.9396747619821925e-05,
|
| 2077 |
+
"loss": 0.2247,
|
| 2078 |
+
"step": 1460
|
| 2079 |
+
},
|
| 2080 |
+
{
|
| 2081 |
+
"epoch": 1.5437302423603794,
|
| 2082 |
+
"grad_norm": 0.451171875,
|
| 2083 |
+
"learning_rate": 4.922921038668035e-05,
|
| 2084 |
+
"loss": 0.2195,
|
| 2085 |
+
"step": 1465
|
| 2086 |
+
},
|
| 2087 |
+
{
|
| 2088 |
+
"epoch": 1.5489989462592202,
|
| 2089 |
+
"grad_norm": 0.41796875,
|
| 2090 |
+
"learning_rate": 4.906169720513176e-05,
|
| 2091 |
+
"loss": 0.222,
|
| 2092 |
+
"step": 1470
|
| 2093 |
+
},
|
| 2094 |
+
{
|
| 2095 |
+
"epoch": 1.554267650158061,
|
| 2096 |
+
"grad_norm": 0.431640625,
|
| 2097 |
+
"learning_rate": 4.8894213302230055e-05,
|
| 2098 |
+
"loss": 0.2275,
|
| 2099 |
+
"step": 1475
|
| 2100 |
+
},
|
| 2101 |
+
{
|
| 2102 |
+
"epoch": 1.559536354056902,
|
| 2103 |
+
"grad_norm": 0.421875,
|
| 2104 |
+
"learning_rate": 4.8726763904115556e-05,
|
| 2105 |
+
"loss": 0.2193,
|
| 2106 |
+
"step": 1480
|
| 2107 |
+
},
|
| 2108 |
+
{
|
| 2109 |
+
"epoch": 1.5648050579557429,
|
| 2110 |
+
"grad_norm": 0.44140625,
|
| 2111 |
+
"learning_rate": 4.8559354235851854e-05,
|
| 2112 |
+
"loss": 0.2289,
|
| 2113 |
+
"step": 1485
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 1.5700737618545837,
|
| 2117 |
+
"grad_norm": 0.41796875,
|
| 2118 |
+
"learning_rate": 4.839198952126288e-05,
|
| 2119 |
+
"loss": 0.2217,
|
| 2120 |
+
"step": 1490
|
| 2121 |
+
},
|
| 2122 |
+
{
|
| 2123 |
+
"epoch": 1.5753424657534247,
|
| 2124 |
+
"grad_norm": 0.435546875,
|
| 2125 |
+
"learning_rate": 4.8224674982769796e-05,
|
| 2126 |
+
"loss": 0.2146,
|
| 2127 |
+
"step": 1495
|
| 2128 |
+
},
|
| 2129 |
+
{
|
| 2130 |
+
"epoch": 1.5806111696522656,
|
| 2131 |
+
"grad_norm": 0.421875,
|
| 2132 |
+
"learning_rate": 4.805741584122808e-05,
|
| 2133 |
+
"loss": 0.2147,
|
| 2134 |
+
"step": 1500
|
| 2135 |
+
},
|
| 2136 |
+
{
|
| 2137 |
+
"epoch": 1.5806111696522656,
|
| 2138 |
+
"eval_loss": 0.29037579894065857,
|
| 2139 |
+
"eval_runtime": 2.4367,
|
| 2140 |
+
"eval_samples_per_second": 19.289,
|
| 2141 |
+
"eval_steps_per_second": 19.289,
|
| 2142 |
+
"step": 1500
|
| 2143 |
+
},
|
| 2144 |
+
{
|
| 2145 |
+
"epoch": 1.5858798735511064,
|
| 2146 |
+
"grad_norm": 0.435546875,
|
| 2147 |
+
"learning_rate": 4.78902173157646e-05,
|
| 2148 |
+
"loss": 0.2182,
|
| 2149 |
+
"step": 1505
|
| 2150 |
+
},
|
| 2151 |
+
{
|
| 2152 |
+
"epoch": 1.5911485774499474,
|
| 2153 |
+
"grad_norm": 0.423828125,
|
| 2154 |
+
"learning_rate": 4.772308462361483e-05,
|
| 2155 |
+
"loss": 0.2202,
|
| 2156 |
+
"step": 1510
|
| 2157 |
+
},
|
| 2158 |
+
{
|
| 2159 |
+
"epoch": 1.5964172813487882,
|
| 2160 |
+
"grad_norm": 0.4296875,
|
| 2161 |
+
"learning_rate": 4.7556022979959925e-05,
|
| 2162 |
+
"loss": 0.2248,
|
| 2163 |
+
"step": 1515
|
| 2164 |
+
},
|
| 2165 |
+
{
|
| 2166 |
+
"epoch": 1.601685985247629,
|
| 2167 |
+
"grad_norm": 0.41796875,
|
| 2168 |
+
"learning_rate": 4.7389037597764084e-05,
|
| 2169 |
+
"loss": 0.2185,
|
| 2170 |
+
"step": 1520
|
| 2171 |
+
},
|
| 2172 |
+
{
|
| 2173 |
+
"epoch": 1.60695468914647,
|
| 2174 |
+
"grad_norm": 0.412109375,
|
| 2175 |
+
"learning_rate": 4.7222133687611846e-05,
|
| 2176 |
+
"loss": 0.2143,
|
| 2177 |
+
"step": 1525
|
| 2178 |
+
},
|
| 2179 |
+
{
|
| 2180 |
+
"epoch": 1.6122233930453107,
|
| 2181 |
+
"grad_norm": 0.427734375,
|
| 2182 |
+
"learning_rate": 4.705531645754552e-05,
|
| 2183 |
+
"loss": 0.228,
|
| 2184 |
+
"step": 1530
|
| 2185 |
+
},
|
| 2186 |
+
{
|
| 2187 |
+
"epoch": 1.6174920969441517,
|
| 2188 |
+
"grad_norm": 0.427734375,
|
| 2189 |
+
"learning_rate": 4.688859111290268e-05,
|
| 2190 |
+
"loss": 0.2205,
|
| 2191 |
+
"step": 1535
|
| 2192 |
+
},
|
| 2193 |
+
{
|
| 2194 |
+
"epoch": 1.6227608008429928,
|
| 2195 |
+
"grad_norm": 0.435546875,
|
| 2196 |
+
"learning_rate": 4.672196285615367e-05,
|
| 2197 |
+
"loss": 0.2179,
|
| 2198 |
+
"step": 1540
|
| 2199 |
+
},
|
| 2200 |
+
{
|
| 2201 |
+
"epoch": 1.6280295047418334,
|
| 2202 |
+
"grad_norm": 0.421875,
|
| 2203 |
+
"learning_rate": 4.655543688673936e-05,
|
| 2204 |
+
"loss": 0.2252,
|
| 2205 |
+
"step": 1545
|
| 2206 |
+
},
|
| 2207 |
+
{
|
| 2208 |
+
"epoch": 1.6332982086406744,
|
| 2209 |
+
"grad_norm": 0.419921875,
|
| 2210 |
+
"learning_rate": 4.638901840090886e-05,
|
| 2211 |
+
"loss": 0.2207,
|
| 2212 |
+
"step": 1550
|
| 2213 |
+
},
|
| 2214 |
+
{
|
| 2215 |
+
"epoch": 1.6385669125395153,
|
| 2216 |
+
"grad_norm": 0.42578125,
|
| 2217 |
+
"learning_rate": 4.6222712591557375e-05,
|
| 2218 |
+
"loss": 0.2213,
|
| 2219 |
+
"step": 1555
|
| 2220 |
+
},
|
| 2221 |
+
{
|
| 2222 |
+
"epoch": 1.643835616438356,
|
| 2223 |
+
"grad_norm": 0.427734375,
|
| 2224 |
+
"learning_rate": 4.6056524648064163e-05,
|
| 2225 |
+
"loss": 0.2202,
|
| 2226 |
+
"step": 1560
|
| 2227 |
+
},
|
| 2228 |
+
{
|
| 2229 |
+
"epoch": 1.6491043203371971,
|
| 2230 |
+
"grad_norm": 0.419921875,
|
| 2231 |
+
"learning_rate": 4.589045975613062e-05,
|
| 2232 |
+
"loss": 0.2162,
|
| 2233 |
+
"step": 1565
|
| 2234 |
+
},
|
| 2235 |
+
{
|
| 2236 |
+
"epoch": 1.654373024236038,
|
| 2237 |
+
"grad_norm": 0.412109375,
|
| 2238 |
+
"learning_rate": 4.572452309761847e-05,
|
| 2239 |
+
"loss": 0.2169,
|
| 2240 |
+
"step": 1570
|
| 2241 |
+
},
|
| 2242 |
+
{
|
| 2243 |
+
"epoch": 1.6596417281348788,
|
| 2244 |
+
"grad_norm": 0.439453125,
|
| 2245 |
+
"learning_rate": 4.555871985038805e-05,
|
| 2246 |
+
"loss": 0.2237,
|
| 2247 |
+
"step": 1575
|
| 2248 |
+
},
|
| 2249 |
+
{
|
| 2250 |
+
"epoch": 1.6649104320337198,
|
| 2251 |
+
"grad_norm": 0.427734375,
|
| 2252 |
+
"learning_rate": 4.539305518813677e-05,
|
| 2253 |
+
"loss": 0.2207,
|
| 2254 |
+
"step": 1580
|
| 2255 |
+
},
|
| 2256 |
+
{
|
| 2257 |
+
"epoch": 1.6701791359325606,
|
| 2258 |
+
"grad_norm": 0.4140625,
|
| 2259 |
+
"learning_rate": 4.5227534280237645e-05,
|
| 2260 |
+
"loss": 0.2167,
|
| 2261 |
+
"step": 1585
|
| 2262 |
+
},
|
| 2263 |
+
{
|
| 2264 |
+
"epoch": 1.6754478398314014,
|
| 2265 |
+
"grad_norm": 0.423828125,
|
| 2266 |
+
"learning_rate": 4.506216229157797e-05,
|
| 2267 |
+
"loss": 0.2191,
|
| 2268 |
+
"step": 1590
|
| 2269 |
+
},
|
| 2270 |
+
{
|
| 2271 |
+
"epoch": 1.6807165437302425,
|
| 2272 |
+
"grad_norm": 0.423828125,
|
| 2273 |
+
"learning_rate": 4.489694438239827e-05,
|
| 2274 |
+
"loss": 0.2171,
|
| 2275 |
+
"step": 1595
|
| 2276 |
+
},
|
| 2277 |
+
{
|
| 2278 |
+
"epoch": 1.685985247629083,
|
| 2279 |
+
"grad_norm": 0.416015625,
|
| 2280 |
+
"learning_rate": 4.4731885708131135e-05,
|
| 2281 |
+
"loss": 0.2191,
|
| 2282 |
+
"step": 1600
|
| 2283 |
+
},
|
| 2284 |
+
{
|
| 2285 |
+
"epoch": 1.6912539515279241,
|
| 2286 |
+
"grad_norm": 0.427734375,
|
| 2287 |
+
"learning_rate": 4.456699141924041e-05,
|
| 2288 |
+
"loss": 0.2195,
|
| 2289 |
+
"step": 1605
|
| 2290 |
+
},
|
| 2291 |
+
{
|
| 2292 |
+
"epoch": 1.6965226554267652,
|
| 2293 |
+
"grad_norm": 0.419921875,
|
| 2294 |
+
"learning_rate": 4.4402266661060535e-05,
|
| 2295 |
+
"loss": 0.2168,
|
| 2296 |
+
"step": 1610
|
| 2297 |
+
},
|
| 2298 |
+
{
|
| 2299 |
+
"epoch": 1.7017913593256058,
|
| 2300 |
+
"grad_norm": 0.42578125,
|
| 2301 |
+
"learning_rate": 4.4237716573635895e-05,
|
| 2302 |
+
"loss": 0.22,
|
| 2303 |
+
"step": 1615
|
| 2304 |
+
},
|
| 2305 |
+
{
|
| 2306 |
+
"epoch": 1.7070600632244468,
|
| 2307 |
+
"grad_norm": 0.4140625,
|
| 2308 |
+
"learning_rate": 4.40733462915605e-05,
|
| 2309 |
+
"loss": 0.2103,
|
| 2310 |
+
"step": 1620
|
| 2311 |
+
},
|
| 2312 |
+
{
|
| 2313 |
+
"epoch": 1.7123287671232876,
|
| 2314 |
+
"grad_norm": 0.453125,
|
| 2315 |
+
"learning_rate": 4.390916094381774e-05,
|
| 2316 |
+
"loss": 0.2154,
|
| 2317 |
+
"step": 1625
|
| 2318 |
+
},
|
| 2319 |
+
{
|
| 2320 |
+
"epoch": 1.7175974710221285,
|
| 2321 |
+
"grad_norm": 0.45703125,
|
| 2322 |
+
"learning_rate": 4.374516565362034e-05,
|
| 2323 |
+
"loss": 0.223,
|
| 2324 |
+
"step": 1630
|
| 2325 |
+
},
|
| 2326 |
+
{
|
| 2327 |
+
"epoch": 1.7228661749209695,
|
| 2328 |
+
"grad_norm": 0.419921875,
|
| 2329 |
+
"learning_rate": 4.35813655382505e-05,
|
| 2330 |
+
"loss": 0.2193,
|
| 2331 |
+
"step": 1635
|
| 2332 |
+
},
|
| 2333 |
+
{
|
| 2334 |
+
"epoch": 1.7281348788198103,
|
| 2335 |
+
"grad_norm": 0.4453125,
|
| 2336 |
+
"learning_rate": 4.341776570890024e-05,
|
| 2337 |
+
"loss": 0.213,
|
| 2338 |
+
"step": 1640
|
| 2339 |
+
},
|
| 2340 |
+
{
|
| 2341 |
+
"epoch": 1.7334035827186511,
|
| 2342 |
+
"grad_norm": 0.41796875,
|
| 2343 |
+
"learning_rate": 4.325437127051184e-05,
|
| 2344 |
+
"loss": 0.219,
|
| 2345 |
+
"step": 1645
|
| 2346 |
+
},
|
| 2347 |
+
{
|
| 2348 |
+
"epoch": 1.7386722866174922,
|
| 2349 |
+
"grad_norm": 0.419921875,
|
| 2350 |
+
"learning_rate": 4.309118732161865e-05,
|
| 2351 |
+
"loss": 0.2188,
|
| 2352 |
+
"step": 1650
|
| 2353 |
+
},
|
| 2354 |
+
{
|
| 2355 |
+
"epoch": 1.743940990516333,
|
| 2356 |
+
"grad_norm": 0.427734375,
|
| 2357 |
+
"learning_rate": 4.292821895418588e-05,
|
| 2358 |
+
"loss": 0.2181,
|
| 2359 |
+
"step": 1655
|
| 2360 |
+
},
|
| 2361 |
+
{
|
| 2362 |
+
"epoch": 1.7492096944151738,
|
| 2363 |
+
"grad_norm": 0.42578125,
|
| 2364 |
+
"learning_rate": 4.2765471253451824e-05,
|
| 2365 |
+
"loss": 0.2168,
|
| 2366 |
+
"step": 1660
|
| 2367 |
+
},
|
| 2368 |
+
{
|
| 2369 |
+
"epoch": 1.7544783983140149,
|
| 2370 |
+
"grad_norm": 0.4296875,
|
| 2371 |
+
"learning_rate": 4.260294929776911e-05,
|
| 2372 |
+
"loss": 0.2231,
|
| 2373 |
+
"step": 1665
|
| 2374 |
+
},
|
| 2375 |
+
{
|
| 2376 |
+
"epoch": 1.7597471022128557,
|
| 2377 |
+
"grad_norm": 0.41015625,
|
| 2378 |
+
"learning_rate": 4.244065815844624e-05,
|
| 2379 |
+
"loss": 0.2084,
|
| 2380 |
+
"step": 1670
|
| 2381 |
+
},
|
| 2382 |
+
{
|
| 2383 |
+
"epoch": 1.7650158061116965,
|
| 2384 |
+
"grad_norm": 0.421875,
|
| 2385 |
+
"learning_rate": 4.227860289958938e-05,
|
| 2386 |
+
"loss": 0.2095,
|
| 2387 |
+
"step": 1675
|
| 2388 |
+
},
|
| 2389 |
+
{
|
| 2390 |
+
"epoch": 1.7702845100105375,
|
| 2391 |
+
"grad_norm": 0.431640625,
|
| 2392 |
+
"learning_rate": 4.211678857794432e-05,
|
| 2393 |
+
"loss": 0.2118,
|
| 2394 |
+
"step": 1680
|
| 2395 |
+
},
|
| 2396 |
+
{
|
| 2397 |
+
"epoch": 1.7755532139093781,
|
| 2398 |
+
"grad_norm": 0.44140625,
|
| 2399 |
+
"learning_rate": 4.1955220242738666e-05,
|
| 2400 |
+
"loss": 0.2131,
|
| 2401 |
+
"step": 1685
|
| 2402 |
+
},
|
| 2403 |
+
{
|
| 2404 |
+
"epoch": 1.7808219178082192,
|
| 2405 |
+
"grad_norm": 0.427734375,
|
| 2406 |
+
"learning_rate": 4.1793902935524314e-05,
|
| 2407 |
+
"loss": 0.2141,
|
| 2408 |
+
"step": 1690
|
| 2409 |
+
},
|
| 2410 |
+
{
|
| 2411 |
+
"epoch": 1.78609062170706,
|
| 2412 |
+
"grad_norm": 0.421875,
|
| 2413 |
+
"learning_rate": 4.163284169002013e-05,
|
| 2414 |
+
"loss": 0.2102,
|
| 2415 |
+
"step": 1695
|
| 2416 |
+
},
|
| 2417 |
+
{
|
| 2418 |
+
"epoch": 1.7913593256059008,
|
| 2419 |
+
"grad_norm": 0.435546875,
|
| 2420 |
+
"learning_rate": 4.147204153195486e-05,
|
| 2421 |
+
"loss": 0.2179,
|
| 2422 |
+
"step": 1700
|
| 2423 |
+
},
|
| 2424 |
+
{
|
| 2425 |
+
"epoch": 1.7966280295047419,
|
| 2426 |
+
"grad_norm": 0.435546875,
|
| 2427 |
+
"learning_rate": 4.1311507478910346e-05,
|
| 2428 |
+
"loss": 0.2133,
|
| 2429 |
+
"step": 1705
|
| 2430 |
+
},
|
| 2431 |
+
{
|
| 2432 |
+
"epoch": 1.8018967334035827,
|
| 2433 |
+
"grad_norm": 0.453125,
|
| 2434 |
+
"learning_rate": 4.11512445401649e-05,
|
| 2435 |
+
"loss": 0.2259,
|
| 2436 |
+
"step": 1710
|
| 2437 |
+
},
|
| 2438 |
+
{
|
| 2439 |
+
"epoch": 1.8071654373024235,
|
| 2440 |
+
"grad_norm": 0.421875,
|
| 2441 |
+
"learning_rate": 4.099125771653703e-05,
|
| 2442 |
+
"loss": 0.2247,
|
| 2443 |
+
"step": 1715
|
| 2444 |
+
},
|
| 2445 |
+
{
|
| 2446 |
+
"epoch": 1.8124341412012646,
|
| 2447 |
+
"grad_norm": 0.423828125,
|
| 2448 |
+
"learning_rate": 4.083155200022942e-05,
|
| 2449 |
+
"loss": 0.2104,
|
| 2450 |
+
"step": 1720
|
| 2451 |
+
},
|
| 2452 |
+
{
|
| 2453 |
+
"epoch": 1.8177028451001054,
|
| 2454 |
+
"grad_norm": 0.416015625,
|
| 2455 |
+
"learning_rate": 4.067213237467312e-05,
|
| 2456 |
+
"loss": 0.2152,
|
| 2457 |
+
"step": 1725
|
| 2458 |
+
},
|
| 2459 |
+
{
|
| 2460 |
+
"epoch": 1.8229715489989462,
|
| 2461 |
+
"grad_norm": 0.431640625,
|
| 2462 |
+
"learning_rate": 4.051300381437201e-05,
|
| 2463 |
+
"loss": 0.2213,
|
| 2464 |
+
"step": 1730
|
| 2465 |
+
},
|
| 2466 |
+
{
|
| 2467 |
+
"epoch": 1.8282402528977872,
|
| 2468 |
+
"grad_norm": 0.408203125,
|
| 2469 |
+
"learning_rate": 4.035417128474769e-05,
|
| 2470 |
+
"loss": 0.2156,
|
| 2471 |
+
"step": 1735
|
| 2472 |
+
},
|
| 2473 |
+
{
|
| 2474 |
+
"epoch": 1.833508956796628,
|
| 2475 |
+
"grad_norm": 0.423828125,
|
| 2476 |
+
"learning_rate": 4.019563974198439e-05,
|
| 2477 |
+
"loss": 0.2119,
|
| 2478 |
+
"step": 1740
|
| 2479 |
+
},
|
| 2480 |
+
{
|
| 2481 |
+
"epoch": 1.8387776606954689,
|
| 2482 |
+
"grad_norm": 0.419921875,
|
| 2483 |
+
"learning_rate": 4.0037414132874454e-05,
|
| 2484 |
+
"loss": 0.2149,
|
| 2485 |
+
"step": 1745
|
| 2486 |
+
},
|
| 2487 |
+
{
|
| 2488 |
+
"epoch": 1.84404636459431,
|
| 2489 |
+
"grad_norm": 0.404296875,
|
| 2490 |
+
"learning_rate": 3.98794993946639e-05,
|
| 2491 |
+
"loss": 0.207,
|
| 2492 |
+
"step": 1750
|
| 2493 |
+
},
|
| 2494 |
+
{
|
| 2495 |
+
"epoch": 1.8493150684931505,
|
| 2496 |
+
"grad_norm": 0.423828125,
|
| 2497 |
+
"learning_rate": 3.972190045489838e-05,
|
| 2498 |
+
"loss": 0.207,
|
| 2499 |
+
"step": 1755
|
| 2500 |
+
},
|
| 2501 |
+
{
|
| 2502 |
+
"epoch": 1.8545837723919916,
|
| 2503 |
+
"grad_norm": 0.412109375,
|
| 2504 |
+
"learning_rate": 3.956462223126941e-05,
|
| 2505 |
+
"loss": 0.2176,
|
| 2506 |
+
"step": 1760
|
| 2507 |
+
},
|
| 2508 |
+
{
|
| 2509 |
+
"epoch": 1.8598524762908326,
|
| 2510 |
+
"grad_norm": 0.419921875,
|
| 2511 |
+
"learning_rate": 3.940766963146097e-05,
|
| 2512 |
+
"loss": 0.2195,
|
| 2513 |
+
"step": 1765
|
| 2514 |
+
},
|
| 2515 |
+
{
|
| 2516 |
+
"epoch": 1.8651211801896732,
|
| 2517 |
+
"grad_norm": 0.419921875,
|
| 2518 |
+
"learning_rate": 3.9251047552996304e-05,
|
| 2519 |
+
"loss": 0.2203,
|
| 2520 |
+
"step": 1770
|
| 2521 |
+
},
|
| 2522 |
+
{
|
| 2523 |
+
"epoch": 1.8703898840885143,
|
| 2524 |
+
"grad_norm": 0.43359375,
|
| 2525 |
+
"learning_rate": 3.9094760883085096e-05,
|
| 2526 |
+
"loss": 0.2217,
|
| 2527 |
+
"step": 1775
|
| 2528 |
+
},
|
| 2529 |
+
{
|
| 2530 |
+
"epoch": 1.875658587987355,
|
| 2531 |
+
"grad_norm": 0.404296875,
|
| 2532 |
+
"learning_rate": 3.8938814498471055e-05,
|
| 2533 |
+
"loss": 0.2036,
|
| 2534 |
+
"step": 1780
|
| 2535 |
+
},
|
| 2536 |
+
{
|
| 2537 |
+
"epoch": 1.880927291886196,
|
| 2538 |
+
"grad_norm": 0.41796875,
|
| 2539 |
+
"learning_rate": 3.8783213265279634e-05,
|
| 2540 |
+
"loss": 0.2165,
|
| 2541 |
+
"step": 1785
|
| 2542 |
+
},
|
| 2543 |
+
{
|
| 2544 |
+
"epoch": 1.886195995785037,
|
| 2545 |
+
"grad_norm": 0.41796875,
|
| 2546 |
+
"learning_rate": 3.8627962038866255e-05,
|
| 2547 |
+
"loss": 0.2108,
|
| 2548 |
+
"step": 1790
|
| 2549 |
+
},
|
| 2550 |
+
{
|
| 2551 |
+
"epoch": 1.8914646996838778,
|
| 2552 |
+
"grad_norm": 0.43359375,
|
| 2553 |
+
"learning_rate": 3.84730656636648e-05,
|
| 2554 |
+
"loss": 0.2106,
|
| 2555 |
+
"step": 1795
|
| 2556 |
+
},
|
| 2557 |
+
{
|
| 2558 |
+
"epoch": 1.8967334035827186,
|
| 2559 |
+
"grad_norm": 0.41796875,
|
| 2560 |
+
"learning_rate": 3.8318528973036395e-05,
|
| 2561 |
+
"loss": 0.2144,
|
| 2562 |
+
"step": 1800
|
| 2563 |
+
},
|
| 2564 |
+
{
|
| 2565 |
+
"epoch": 1.9020021074815596,
|
| 2566 |
+
"grad_norm": 0.41796875,
|
| 2567 |
+
"learning_rate": 3.816435678911868e-05,
|
| 2568 |
+
"loss": 0.2099,
|
| 2569 |
+
"step": 1805
|
| 2570 |
+
},
|
| 2571 |
+
{
|
| 2572 |
+
"epoch": 1.9072708113804004,
|
| 2573 |
+
"grad_norm": 0.408203125,
|
| 2574 |
+
"learning_rate": 3.801055392267523e-05,
|
| 2575 |
+
"loss": 0.2168,
|
| 2576 |
+
"step": 1810
|
| 2577 |
+
},
|
| 2578 |
+
{
|
| 2579 |
+
"epoch": 1.9125395152792413,
|
| 2580 |
+
"grad_norm": 0.408203125,
|
| 2581 |
+
"learning_rate": 3.785712517294552e-05,
|
| 2582 |
+
"loss": 0.2062,
|
| 2583 |
+
"step": 1815
|
| 2584 |
+
},
|
| 2585 |
+
{
|
| 2586 |
+
"epoch": 1.9178082191780823,
|
| 2587 |
+
"grad_norm": 0.439453125,
|
| 2588 |
+
"learning_rate": 3.770407532749519e-05,
|
| 2589 |
+
"loss": 0.2215,
|
| 2590 |
+
"step": 1820
|
| 2591 |
+
},
|
| 2592 |
+
{
|
| 2593 |
+
"epoch": 1.9230769230769231,
|
| 2594 |
+
"grad_norm": 0.416015625,
|
| 2595 |
+
"learning_rate": 3.755140916206654e-05,
|
| 2596 |
+
"loss": 0.2111,
|
| 2597 |
+
"step": 1825
|
| 2598 |
+
},
|
| 2599 |
+
{
|
| 2600 |
+
"epoch": 1.928345626975764,
|
| 2601 |
+
"grad_norm": 0.421875,
|
| 2602 |
+
"learning_rate": 3.739913144042963e-05,
|
| 2603 |
+
"loss": 0.2109,
|
| 2604 |
+
"step": 1830
|
| 2605 |
+
},
|
| 2606 |
+
{
|
| 2607 |
+
"epoch": 1.933614330874605,
|
| 2608 |
+
"grad_norm": 0.416015625,
|
| 2609 |
+
"learning_rate": 3.7247246914233584e-05,
|
| 2610 |
+
"loss": 0.2133,
|
| 2611 |
+
"step": 1835
|
| 2612 |
+
},
|
| 2613 |
+
{
|
| 2614 |
+
"epoch": 1.9388830347734456,
|
| 2615 |
+
"grad_norm": 0.42578125,
|
| 2616 |
+
"learning_rate": 3.709576032285829e-05,
|
| 2617 |
+
"loss": 0.2128,
|
| 2618 |
+
"step": 1840
|
| 2619 |
+
},
|
| 2620 |
+
{
|
| 2621 |
+
"epoch": 1.9441517386722866,
|
| 2622 |
+
"grad_norm": 0.42578125,
|
| 2623 |
+
"learning_rate": 3.694467639326656e-05,
|
| 2624 |
+
"loss": 0.2161,
|
| 2625 |
+
"step": 1845
|
| 2626 |
+
},
|
| 2627 |
+
{
|
| 2628 |
+
"epoch": 1.9494204425711275,
|
| 2629 |
+
"grad_norm": 0.43359375,
|
| 2630 |
+
"learning_rate": 3.679399983985663e-05,
|
| 2631 |
+
"loss": 0.2215,
|
| 2632 |
+
"step": 1850
|
| 2633 |
+
},
|
| 2634 |
+
{
|
| 2635 |
+
"epoch": 1.9546891464699683,
|
| 2636 |
+
"grad_norm": 0.412109375,
|
| 2637 |
+
"learning_rate": 3.6643735364314995e-05,
|
| 2638 |
+
"loss": 0.2111,
|
| 2639 |
+
"step": 1855
|
| 2640 |
+
},
|
| 2641 |
+
{
|
| 2642 |
+
"epoch": 1.9599578503688093,
|
| 2643 |
+
"grad_norm": 0.41015625,
|
| 2644 |
+
"learning_rate": 3.6493887655469796e-05,
|
| 2645 |
+
"loss": 0.2141,
|
| 2646 |
+
"step": 1860
|
| 2647 |
+
},
|
| 2648 |
+
{
|
| 2649 |
+
"epoch": 1.9652265542676501,
|
| 2650 |
+
"grad_norm": 0.4375,
|
| 2651 |
+
"learning_rate": 3.63444613891444e-05,
|
| 2652 |
+
"loss": 0.2102,
|
| 2653 |
+
"step": 1865
|
| 2654 |
+
},
|
| 2655 |
+
{
|
| 2656 |
+
"epoch": 1.970495258166491,
|
| 2657 |
+
"grad_norm": 0.412109375,
|
| 2658 |
+
"learning_rate": 3.619546122801158e-05,
|
| 2659 |
+
"loss": 0.204,
|
| 2660 |
+
"step": 1870
|
| 2661 |
+
},
|
| 2662 |
+
{
|
| 2663 |
+
"epoch": 1.975763962065332,
|
| 2664 |
+
"grad_norm": 0.419921875,
|
| 2665 |
+
"learning_rate": 3.604689182144798e-05,
|
| 2666 |
+
"loss": 0.2057,
|
| 2667 |
+
"step": 1875
|
| 2668 |
+
},
|
| 2669 |
+
{
|
| 2670 |
+
"epoch": 1.9810326659641728,
|
| 2671 |
+
"grad_norm": 0.42578125,
|
| 2672 |
+
"learning_rate": 3.589875780538906e-05,
|
| 2673 |
+
"loss": 0.2107,
|
| 2674 |
+
"step": 1880
|
| 2675 |
+
},
|
| 2676 |
+
{
|
| 2677 |
+
"epoch": 1.9863013698630136,
|
| 2678 |
+
"grad_norm": 0.4375,
|
| 2679 |
+
"learning_rate": 3.575106380218442e-05,
|
| 2680 |
+
"loss": 0.2202,
|
| 2681 |
+
"step": 1885
|
| 2682 |
+
},
|
| 2683 |
+
{
|
| 2684 |
+
"epoch": 1.9915700737618547,
|
| 2685 |
+
"grad_norm": 0.4140625,
|
| 2686 |
+
"learning_rate": 3.5603814420453566e-05,
|
| 2687 |
+
"loss": 0.2075,
|
| 2688 |
+
"step": 1890
|
| 2689 |
+
},
|
| 2690 |
+
{
|
| 2691 |
+
"epoch": 1.9968387776606955,
|
| 2692 |
+
"grad_norm": 0.416015625,
|
| 2693 |
+
"learning_rate": 3.5457014254942126e-05,
|
| 2694 |
+
"loss": 0.2163,
|
| 2695 |
+
"step": 1895
|
| 2696 |
+
},
|
| 2697 |
+
{
|
| 2698 |
+
"epoch": 1.9978925184404637,
|
| 2699 |
+
"eval_loss": 0.28488120436668396,
|
| 2700 |
+
"eval_runtime": 2.5242,
|
| 2701 |
+
"eval_samples_per_second": 18.619,
|
| 2702 |
+
"eval_steps_per_second": 18.619,
|
| 2703 |
+
"step": 1896
|
| 2704 |
+
}
|
| 2705 |
+
],
|
| 2706 |
+
"logging_steps": 5,
|
| 2707 |
+
"max_steps": 2847,
|
| 2708 |
+
"num_input_tokens_seen": 0,
|
| 2709 |
+
"num_train_epochs": 3,
|
| 2710 |
+
"save_steps": 500,
|
| 2711 |
+
"stateful_callbacks": {
|
| 2712 |
+
"TrainerControl": {
|
| 2713 |
+
"args": {
|
| 2714 |
+
"should_epoch_stop": false,
|
| 2715 |
+
"should_evaluate": false,
|
| 2716 |
+
"should_log": false,
|
| 2717 |
+
"should_save": true,
|
| 2718 |
+
"should_training_stop": true
|
| 2719 |
+
},
|
| 2720 |
+
"attributes": {}
|
| 2721 |
+
}
|
| 2722 |
+
},
|
| 2723 |
+
"total_flos": 3.051871750229852e+18,
|
| 2724 |
+
"train_batch_size": 100,
|
| 2725 |
+
"trial_name": null,
|
| 2726 |
+
"trial_params": null
|
| 2727 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:63dcfa6fe661e6522819316451dcbd41131cc208eb88ad00e77b83a795a4ccfe
|
| 3 |
+
size 5624
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|