Upload task output 281ac9fc-853b-477d-8e84-145f9030d732
Browse files- config.json +31 -0
- generation_config.json +7 -0
- loss.txt +1 -0
- model.safetensors +3 -0
- special_tokens_map.json +30 -0
- tokenizer.json +0 -0
- tokenizer.model +3 -0
- tokenizer_config.json +44 -0
- trainer_state.json +2183 -0
- training_args.bin +3 -0
config.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"LlamaForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 1,
|
| 8 |
+
"eos_token_id": 2,
|
| 9 |
+
"head_dim": 64,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 2048,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 5632,
|
| 14 |
+
"max_position_embeddings": 2048,
|
| 15 |
+
"mlp_bias": false,
|
| 16 |
+
"model_type": "llama",
|
| 17 |
+
"num_attention_heads": 32,
|
| 18 |
+
"num_hidden_layers": 22,
|
| 19 |
+
"num_key_value_heads": 4,
|
| 20 |
+
"pad_token_id": 0,
|
| 21 |
+
"pretraining_tp": 1,
|
| 22 |
+
"rms_norm_eps": 1e-05,
|
| 23 |
+
"rope_scaling": null,
|
| 24 |
+
"rope_theta": 10000.0,
|
| 25 |
+
"tie_word_embeddings": false,
|
| 26 |
+
"torch_dtype": "bfloat16",
|
| 27 |
+
"transformers_version": "4.51.3",
|
| 28 |
+
"unsloth_version": "2024.9",
|
| 29 |
+
"use_cache": false,
|
| 30 |
+
"vocab_size": 32000
|
| 31 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 1,
|
| 3 |
+
"eos_token_id": 2,
|
| 4 |
+
"max_length": 2048,
|
| 5 |
+
"pad_token_id": 0,
|
| 6 |
+
"transformers_version": "4.51.3"
|
| 7 |
+
}
|
loss.txt
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
699,0.00020370040147099644
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9fd06729cbdb2859971e6ec0e27c0a73f9d8dc707210dcec797f1a2a62bc29e2
|
| 3 |
+
size 2200119864
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": {
|
| 3 |
+
"content": "<s>",
|
| 4 |
+
"lstrip": false,
|
| 5 |
+
"normalized": false,
|
| 6 |
+
"rstrip": false,
|
| 7 |
+
"single_word": false
|
| 8 |
+
},
|
| 9 |
+
"eos_token": {
|
| 10 |
+
"content": "</s>",
|
| 11 |
+
"lstrip": false,
|
| 12 |
+
"normalized": false,
|
| 13 |
+
"rstrip": false,
|
| 14 |
+
"single_word": false
|
| 15 |
+
},
|
| 16 |
+
"pad_token": {
|
| 17 |
+
"content": "<unk>",
|
| 18 |
+
"lstrip": false,
|
| 19 |
+
"normalized": false,
|
| 20 |
+
"rstrip": false,
|
| 21 |
+
"single_word": false
|
| 22 |
+
},
|
| 23 |
+
"unk_token": {
|
| 24 |
+
"content": "<unk>",
|
| 25 |
+
"lstrip": false,
|
| 26 |
+
"normalized": false,
|
| 27 |
+
"rstrip": false,
|
| 28 |
+
"single_word": false
|
| 29 |
+
}
|
| 30 |
+
}
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer.model
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9e556afd44213b6bd1be2b850ebbbd98f5481437a8021afaf58ee7fb1818d347
|
| 3 |
+
size 499723
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,44 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": true,
|
| 3 |
+
"add_eos_token": false,
|
| 4 |
+
"add_prefix_space": null,
|
| 5 |
+
"added_tokens_decoder": {
|
| 6 |
+
"0": {
|
| 7 |
+
"content": "<unk>",
|
| 8 |
+
"lstrip": false,
|
| 9 |
+
"normalized": false,
|
| 10 |
+
"rstrip": false,
|
| 11 |
+
"single_word": false,
|
| 12 |
+
"special": true
|
| 13 |
+
},
|
| 14 |
+
"1": {
|
| 15 |
+
"content": "<s>",
|
| 16 |
+
"lstrip": false,
|
| 17 |
+
"normalized": false,
|
| 18 |
+
"rstrip": false,
|
| 19 |
+
"single_word": false,
|
| 20 |
+
"special": true
|
| 21 |
+
},
|
| 22 |
+
"2": {
|
| 23 |
+
"content": "</s>",
|
| 24 |
+
"lstrip": false,
|
| 25 |
+
"normalized": false,
|
| 26 |
+
"rstrip": false,
|
| 27 |
+
"single_word": false,
|
| 28 |
+
"special": true
|
| 29 |
+
}
|
| 30 |
+
},
|
| 31 |
+
"bos_token": "<s>",
|
| 32 |
+
"chat_template": "{% for message in messages %}\n{% if message['role'] == 'user' %}\n{{ '<|user|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'system' %}\n{{ '<|system|>\n' + message['content'] + eos_token }}\n{% elif message['role'] == 'assistant' %}\n{{ '<|assistant|>\n' + message['content'] + eos_token }}\n{% endif %}\n{% if loop.last and add_generation_prompt %}\n{{ '<|assistant|>' }}\n{% endif %}\n{% endfor %}",
|
| 33 |
+
"clean_up_tokenization_spaces": false,
|
| 34 |
+
"eos_token": "</s>",
|
| 35 |
+
"extra_special_tokens": {},
|
| 36 |
+
"legacy": false,
|
| 37 |
+
"model_max_length": 2048,
|
| 38 |
+
"pad_token": "<unk>",
|
| 39 |
+
"padding_side": "left",
|
| 40 |
+
"sp_model_kwargs": {},
|
| 41 |
+
"tokenizer_class": "LlamaTokenizer",
|
| 42 |
+
"unk_token": "<unk>",
|
| 43 |
+
"use_default_system_prompt": false
|
| 44 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,2183 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 2.9871794871794872,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 699,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"epoch": 0.021367521367521368,
|
| 14 |
+
"grad_norm": 16.125,
|
| 15 |
+
"learning_rate": 1.7556090538745385e-06,
|
| 16 |
+
"logits/chosen": -3.5722389221191406,
|
| 17 |
+
"logits/rejected": -3.5034377574920654,
|
| 18 |
+
"logps/chosen": -41.095054626464844,
|
| 19 |
+
"logps/rejected": -79.83882141113281,
|
| 20 |
+
"loss": 0.6922,
|
| 21 |
+
"rewards/accuracies": 0.3500000238418579,
|
| 22 |
+
"rewards/chosen": 0.0014523781137540936,
|
| 23 |
+
"rewards/margins": 0.001936142100021243,
|
| 24 |
+
"rewards/rejected": -0.0004837641608901322,
|
| 25 |
+
"step": 5
|
| 26 |
+
},
|
| 27 |
+
{
|
| 28 |
+
"epoch": 0.042735042735042736,
|
| 29 |
+
"grad_norm": 16.75,
|
| 30 |
+
"learning_rate": 3.950120371217711e-06,
|
| 31 |
+
"logits/chosen": -3.586623430252075,
|
| 32 |
+
"logits/rejected": -3.506187915802002,
|
| 33 |
+
"logps/chosen": -40.016441345214844,
|
| 34 |
+
"logps/rejected": -78.24286651611328,
|
| 35 |
+
"loss": 0.6796,
|
| 36 |
+
"rewards/accuracies": 0.824999988079071,
|
| 37 |
+
"rewards/chosen": -0.002693606074899435,
|
| 38 |
+
"rewards/margins": 0.027483653277158737,
|
| 39 |
+
"rewards/rejected": -0.03017725981771946,
|
| 40 |
+
"step": 10
|
| 41 |
+
},
|
| 42 |
+
{
|
| 43 |
+
"epoch": 0.0641025641025641,
|
| 44 |
+
"grad_norm": 14.75,
|
| 45 |
+
"learning_rate": 6.144631688560886e-06,
|
| 46 |
+
"logits/chosen": -3.5991673469543457,
|
| 47 |
+
"logits/rejected": -3.50789213180542,
|
| 48 |
+
"logps/chosen": -39.836097717285156,
|
| 49 |
+
"logps/rejected": -81.42815399169922,
|
| 50 |
+
"loss": 0.6306,
|
| 51 |
+
"rewards/accuracies": 0.9916666746139526,
|
| 52 |
+
"rewards/chosen": -0.006502463016659021,
|
| 53 |
+
"rewards/margins": 0.1313387155532837,
|
| 54 |
+
"rewards/rejected": -0.13784119486808777,
|
| 55 |
+
"step": 15
|
| 56 |
+
},
|
| 57 |
+
{
|
| 58 |
+
"epoch": 0.08547008547008547,
|
| 59 |
+
"grad_norm": 13.625,
|
| 60 |
+
"learning_rate": 8.339143005904057e-06,
|
| 61 |
+
"logits/chosen": -3.5530943870544434,
|
| 62 |
+
"logits/rejected": -3.492051601409912,
|
| 63 |
+
"logps/chosen": -39.75938034057617,
|
| 64 |
+
"logps/rejected": -80.72886657714844,
|
| 65 |
+
"loss": 0.5434,
|
| 66 |
+
"rewards/accuracies": 0.9791666865348816,
|
| 67 |
+
"rewards/chosen": -0.015247734263539314,
|
| 68 |
+
"rewards/margins": 0.33922019600868225,
|
| 69 |
+
"rewards/rejected": -0.3544679284095764,
|
| 70 |
+
"step": 20
|
| 71 |
+
},
|
| 72 |
+
{
|
| 73 |
+
"epoch": 0.10683760683760683,
|
| 74 |
+
"grad_norm": 11.1875,
|
| 75 |
+
"learning_rate": 1.0533654323247232e-05,
|
| 76 |
+
"logits/chosen": -3.512582302093506,
|
| 77 |
+
"logits/rejected": -3.4719951152801514,
|
| 78 |
+
"logps/chosen": -39.996891021728516,
|
| 79 |
+
"logps/rejected": -85.54742431640625,
|
| 80 |
+
"loss": 0.3961,
|
| 81 |
+
"rewards/accuracies": 0.98333340883255,
|
| 82 |
+
"rewards/chosen": -0.04132762551307678,
|
| 83 |
+
"rewards/margins": 0.7773466110229492,
|
| 84 |
+
"rewards/rejected": -0.8186742067337036,
|
| 85 |
+
"step": 25
|
| 86 |
+
},
|
| 87 |
+
{
|
| 88 |
+
"epoch": 0.1282051282051282,
|
| 89 |
+
"grad_norm": 7.5625,
|
| 90 |
+
"learning_rate": 1.2728165640590407e-05,
|
| 91 |
+
"logits/chosen": -3.4401345252990723,
|
| 92 |
+
"logits/rejected": -3.4451375007629395,
|
| 93 |
+
"logps/chosen": -42.438926696777344,
|
| 94 |
+
"logps/rejected": -93.06343078613281,
|
| 95 |
+
"loss": 0.2584,
|
| 96 |
+
"rewards/accuracies": 0.9958332777023315,
|
| 97 |
+
"rewards/chosen": -0.13156814873218536,
|
| 98 |
+
"rewards/margins": 1.4302090406417847,
|
| 99 |
+
"rewards/rejected": -1.561777114868164,
|
| 100 |
+
"step": 30
|
| 101 |
+
},
|
| 102 |
+
{
|
| 103 |
+
"epoch": 0.14957264957264957,
|
| 104 |
+
"grad_norm": 4.28125,
|
| 105 |
+
"learning_rate": 1.4922676957933578e-05,
|
| 106 |
+
"logits/chosen": -3.220163345336914,
|
| 107 |
+
"logits/rejected": -3.2657477855682373,
|
| 108 |
+
"logps/chosen": -44.191307067871094,
|
| 109 |
+
"logps/rejected": -108.76808166503906,
|
| 110 |
+
"loss": 0.1195,
|
| 111 |
+
"rewards/accuracies": 1.0,
|
| 112 |
+
"rewards/chosen": -0.4397502541542053,
|
| 113 |
+
"rewards/margins": 2.7580618858337402,
|
| 114 |
+
"rewards/rejected": -3.19781231880188,
|
| 115 |
+
"step": 35
|
| 116 |
+
},
|
| 117 |
+
{
|
| 118 |
+
"epoch": 0.17094017094017094,
|
| 119 |
+
"grad_norm": 2.0625,
|
| 120 |
+
"learning_rate": 1.5360556888469565e-05,
|
| 121 |
+
"logits/chosen": -2.7749853134155273,
|
| 122 |
+
"logits/rejected": -2.807795286178589,
|
| 123 |
+
"logps/chosen": -52.53513717651367,
|
| 124 |
+
"logps/rejected": -143.27357482910156,
|
| 125 |
+
"loss": 0.0426,
|
| 126 |
+
"rewards/accuracies": 0.9916666746139526,
|
| 127 |
+
"rewards/chosen": -1.2650867700576782,
|
| 128 |
+
"rewards/margins": 5.236789226531982,
|
| 129 |
+
"rewards/rejected": -6.501875877380371,
|
| 130 |
+
"step": 40
|
| 131 |
+
},
|
| 132 |
+
{
|
| 133 |
+
"epoch": 0.19230769230769232,
|
| 134 |
+
"grad_norm": 1.71875,
|
| 135 |
+
"learning_rate": 1.535640428282884e-05,
|
| 136 |
+
"logits/chosen": -2.3302321434020996,
|
| 137 |
+
"logits/rejected": -2.289696216583252,
|
| 138 |
+
"logps/chosen": -51.40728759765625,
|
| 139 |
+
"logps/rejected": -162.12364196777344,
|
| 140 |
+
"loss": 0.0168,
|
| 141 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 142 |
+
"rewards/chosen": -1.279262900352478,
|
| 143 |
+
"rewards/margins": 7.274726867675781,
|
| 144 |
+
"rewards/rejected": -8.553990364074707,
|
| 145 |
+
"step": 45
|
| 146 |
+
},
|
| 147 |
+
{
|
| 148 |
+
"epoch": 0.21367521367521367,
|
| 149 |
+
"grad_norm": 6.5,
|
| 150 |
+
"learning_rate": 1.5349059809872097e-05,
|
| 151 |
+
"logits/chosen": -1.9786951541900635,
|
| 152 |
+
"logits/rejected": -1.9170547723770142,
|
| 153 |
+
"logps/chosen": -57.05157470703125,
|
| 154 |
+
"logps/rejected": -180.14271545410156,
|
| 155 |
+
"loss": 0.0191,
|
| 156 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 157 |
+
"rewards/chosen": -1.664361596107483,
|
| 158 |
+
"rewards/margins": 8.5733003616333,
|
| 159 |
+
"rewards/rejected": -10.237661361694336,
|
| 160 |
+
"step": 50
|
| 161 |
+
},
|
| 162 |
+
{
|
| 163 |
+
"epoch": 0.23504273504273504,
|
| 164 |
+
"grad_norm": 0.67578125,
|
| 165 |
+
"learning_rate": 1.5338527542732884e-05,
|
| 166 |
+
"logits/chosen": -1.976489782333374,
|
| 167 |
+
"logits/rejected": -1.909419059753418,
|
| 168 |
+
"logps/chosen": -58.60235595703125,
|
| 169 |
+
"logps/rejected": -196.4000701904297,
|
| 170 |
+
"loss": 0.0097,
|
| 171 |
+
"rewards/accuracies": 0.9958332777023315,
|
| 172 |
+
"rewards/chosen": -1.8706943988800049,
|
| 173 |
+
"rewards/margins": 9.873316764831543,
|
| 174 |
+
"rewards/rejected": -11.744011878967285,
|
| 175 |
+
"step": 55
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"epoch": 0.2564102564102564,
|
| 179 |
+
"grad_norm": 3.234375,
|
| 180 |
+
"learning_rate": 1.532481332244717e-05,
|
| 181 |
+
"logits/chosen": -1.9295637607574463,
|
| 182 |
+
"logits/rejected": -1.817983627319336,
|
| 183 |
+
"logps/chosen": -57.39198684692383,
|
| 184 |
+
"logps/rejected": -199.2140350341797,
|
| 185 |
+
"loss": 0.0047,
|
| 186 |
+
"rewards/accuracies": 1.0,
|
| 187 |
+
"rewards/chosen": -1.8252334594726562,
|
| 188 |
+
"rewards/margins": 10.293539047241211,
|
| 189 |
+
"rewards/rejected": -12.11877155303955,
|
| 190 |
+
"step": 60
|
| 191 |
+
},
|
| 192 |
+
{
|
| 193 |
+
"epoch": 0.2777777777777778,
|
| 194 |
+
"grad_norm": 9.125,
|
| 195 |
+
"learning_rate": 1.5307924754713968e-05,
|
| 196 |
+
"logits/chosen": -1.9402471780776978,
|
| 197 |
+
"logits/rejected": -1.8663572072982788,
|
| 198 |
+
"logps/chosen": -59.858306884765625,
|
| 199 |
+
"logps/rejected": -196.60189819335938,
|
| 200 |
+
"loss": 0.0108,
|
| 201 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 202 |
+
"rewards/chosen": -1.9207541942596436,
|
| 203 |
+
"rewards/margins": 9.978109359741211,
|
| 204 |
+
"rewards/rejected": -11.89886474609375,
|
| 205 |
+
"step": 65
|
| 206 |
+
},
|
| 207 |
+
{
|
| 208 |
+
"epoch": 0.29914529914529914,
|
| 209 |
+
"grad_norm": 0.298828125,
|
| 210 |
+
"learning_rate": 1.528787120567736e-05,
|
| 211 |
+
"logits/chosen": -2.046313762664795,
|
| 212 |
+
"logits/rejected": -1.9556039571762085,
|
| 213 |
+
"logps/chosen": -54.13606643676758,
|
| 214 |
+
"logps/rejected": -192.9593963623047,
|
| 215 |
+
"loss": 0.0026,
|
| 216 |
+
"rewards/accuracies": 1.0,
|
| 217 |
+
"rewards/chosen": -1.4830687046051025,
|
| 218 |
+
"rewards/margins": 10.119759559631348,
|
| 219 |
+
"rewards/rejected": -11.602827072143555,
|
| 220 |
+
"step": 70
|
| 221 |
+
},
|
| 222 |
+
{
|
| 223 |
+
"epoch": 0.32051282051282054,
|
| 224 |
+
"grad_norm": 0.400390625,
|
| 225 |
+
"learning_rate": 1.526466379673215e-05,
|
| 226 |
+
"logits/chosen": -2.0463297367095947,
|
| 227 |
+
"logits/rejected": -1.9465014934539795,
|
| 228 |
+
"logps/chosen": -53.60888671875,
|
| 229 |
+
"logps/rejected": -200.31202697753906,
|
| 230 |
+
"loss": 0.0074,
|
| 231 |
+
"rewards/accuracies": 1.0,
|
| 232 |
+
"rewards/chosen": -1.4517700672149658,
|
| 233 |
+
"rewards/margins": 10.894501686096191,
|
| 234 |
+
"rewards/rejected": -12.346272468566895,
|
| 235 |
+
"step": 75
|
| 236 |
+
},
|
| 237 |
+
{
|
| 238 |
+
"epoch": 0.3418803418803419,
|
| 239 |
+
"grad_norm": 14.0,
|
| 240 |
+
"learning_rate": 1.5238315398356126e-05,
|
| 241 |
+
"logits/chosen": -1.9893850088119507,
|
| 242 |
+
"logits/rejected": -1.8696527481079102,
|
| 243 |
+
"logps/chosen": -55.793907165527344,
|
| 244 |
+
"logps/rejected": -205.0140838623047,
|
| 245 |
+
"loss": 0.0099,
|
| 246 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 247 |
+
"rewards/chosen": -1.6301085948944092,
|
| 248 |
+
"rewards/margins": 11.256315231323242,
|
| 249 |
+
"rewards/rejected": -12.88642406463623,
|
| 250 |
+
"step": 80
|
| 251 |
+
},
|
| 252 |
+
{
|
| 253 |
+
"epoch": 0.36324786324786323,
|
| 254 |
+
"grad_norm": 0.040283203125,
|
| 255 |
+
"learning_rate": 1.5208840622972272e-05,
|
| 256 |
+
"logits/chosen": -1.9942238330841064,
|
| 257 |
+
"logits/rejected": -1.8389371633529663,
|
| 258 |
+
"logps/chosen": -60.305450439453125,
|
| 259 |
+
"logps/rejected": -209.89016723632812,
|
| 260 |
+
"loss": 0.0228,
|
| 261 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 262 |
+
"rewards/chosen": -1.8958022594451904,
|
| 263 |
+
"rewards/margins": 11.26887321472168,
|
| 264 |
+
"rewards/rejected": -13.164674758911133,
|
| 265 |
+
"step": 85
|
| 266 |
+
},
|
| 267 |
+
{
|
| 268 |
+
"epoch": 0.38461538461538464,
|
| 269 |
+
"grad_norm": 0.68359375,
|
| 270 |
+
"learning_rate": 1.5176255816844948e-05,
|
| 271 |
+
"logits/chosen": -1.958398461341858,
|
| 272 |
+
"logits/rejected": -1.7746648788452148,
|
| 273 |
+
"logps/chosen": -53.41706466674805,
|
| 274 |
+
"logps/rejected": -208.37869262695312,
|
| 275 |
+
"loss": 0.0067,
|
| 276 |
+
"rewards/accuracies": 1.0,
|
| 277 |
+
"rewards/chosen": -1.439720869064331,
|
| 278 |
+
"rewards/margins": 11.646267890930176,
|
| 279 |
+
"rewards/rejected": -13.08598804473877,
|
| 280 |
+
"step": 90
|
| 281 |
+
},
|
| 282 |
+
{
|
| 283 |
+
"epoch": 0.405982905982906,
|
| 284 |
+
"grad_norm": 0.67578125,
|
| 285 |
+
"learning_rate": 1.5140579051014502e-05,
|
| 286 |
+
"logits/chosen": -1.9176127910614014,
|
| 287 |
+
"logits/rejected": -1.6993322372436523,
|
| 288 |
+
"logps/chosen": -57.2091178894043,
|
| 289 |
+
"logps/rejected": -224.92782592773438,
|
| 290 |
+
"loss": 0.0026,
|
| 291 |
+
"rewards/accuracies": 1.0,
|
| 292 |
+
"rewards/chosen": -1.716392159461975,
|
| 293 |
+
"rewards/margins": 12.696627616882324,
|
| 294 |
+
"rewards/rejected": -14.413020133972168,
|
| 295 |
+
"step": 95
|
| 296 |
+
},
|
| 297 |
+
{
|
| 298 |
+
"epoch": 0.42735042735042733,
|
| 299 |
+
"grad_norm": 0.3203125,
|
| 300 |
+
"learning_rate": 1.5101830111275334e-05,
|
| 301 |
+
"logits/chosen": -1.874871015548706,
|
| 302 |
+
"logits/rejected": -1.682807207107544,
|
| 303 |
+
"logps/chosen": -63.0499267578125,
|
| 304 |
+
"logps/rejected": -215.8572235107422,
|
| 305 |
+
"loss": 0.0118,
|
| 306 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 307 |
+
"rewards/chosen": -2.3144705295562744,
|
| 308 |
+
"rewards/margins": 11.472696304321289,
|
| 309 |
+
"rewards/rejected": -13.787165641784668,
|
| 310 |
+
"step": 100
|
| 311 |
+
},
|
| 312 |
+
{
|
| 313 |
+
"epoch": 0.44871794871794873,
|
| 314 |
+
"grad_norm": 4.21875,
|
| 315 |
+
"learning_rate": 1.5060030487203004e-05,
|
| 316 |
+
"logits/chosen": -1.8294957876205444,
|
| 317 |
+
"logits/rejected": -1.583496332168579,
|
| 318 |
+
"logps/chosen": -67.67689514160156,
|
| 319 |
+
"logps/rejected": -231.562744140625,
|
| 320 |
+
"loss": 0.006,
|
| 321 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 322 |
+
"rewards/chosen": -2.7576889991760254,
|
| 323 |
+
"rewards/margins": 12.489922523498535,
|
| 324 |
+
"rewards/rejected": -15.247611999511719,
|
| 325 |
+
"step": 105
|
| 326 |
+
},
|
| 327 |
+
{
|
| 328 |
+
"epoch": 0.4700854700854701,
|
| 329 |
+
"grad_norm": 0.1015625,
|
| 330 |
+
"learning_rate": 1.501520336023643e-05,
|
| 331 |
+
"logits/chosen": -1.7654807567596436,
|
| 332 |
+
"logits/rejected": -1.4742016792297363,
|
| 333 |
+
"logps/chosen": -66.52511596679688,
|
| 334 |
+
"logps/rejected": -227.6389923095703,
|
| 335 |
+
"loss": 0.0016,
|
| 336 |
+
"rewards/accuracies": 1.0,
|
| 337 |
+
"rewards/chosen": -2.638422966003418,
|
| 338 |
+
"rewards/margins": 12.420553207397461,
|
| 339 |
+
"rewards/rejected": -15.058975219726562,
|
| 340 |
+
"step": 110
|
| 341 |
+
},
|
| 342 |
+
{
|
| 343 |
+
"epoch": 0.49145299145299143,
|
| 344 |
+
"grad_norm": 1.7421875,
|
| 345 |
+
"learning_rate": 1.4967373590821828e-05,
|
| 346 |
+
"logits/chosen": -1.7195736169815063,
|
| 347 |
+
"logits/rejected": -1.4602675437927246,
|
| 348 |
+
"logps/chosen": -63.84660720825195,
|
| 349 |
+
"logps/rejected": -229.90017700195312,
|
| 350 |
+
"loss": 0.0106,
|
| 351 |
+
"rewards/accuracies": 0.9916666746139526,
|
| 352 |
+
"rewards/chosen": -2.3248353004455566,
|
| 353 |
+
"rewards/margins": 12.83845329284668,
|
| 354 |
+
"rewards/rejected": -15.163289070129395,
|
| 355 |
+
"step": 115
|
| 356 |
+
},
|
| 357 |
+
{
|
| 358 |
+
"epoch": 0.5128205128205128,
|
| 359 |
+
"grad_norm": 0.134765625,
|
| 360 |
+
"learning_rate": 1.491656770462546e-05,
|
| 361 |
+
"logits/chosen": -1.617491364479065,
|
| 362 |
+
"logits/rejected": -1.2875694036483765,
|
| 363 |
+
"logps/chosen": -59.133209228515625,
|
| 364 |
+
"logps/rejected": -232.18191528320312,
|
| 365 |
+
"loss": 0.0038,
|
| 366 |
+
"rewards/accuracies": 1.0,
|
| 367 |
+
"rewards/chosen": -1.9275062084197998,
|
| 368 |
+
"rewards/margins": 13.423171997070312,
|
| 369 |
+
"rewards/rejected": -15.350679397583008,
|
| 370 |
+
"step": 120
|
| 371 |
+
},
|
| 372 |
+
{
|
| 373 |
+
"epoch": 0.5341880341880342,
|
| 374 |
+
"grad_norm": 11.0,
|
| 375 |
+
"learning_rate": 1.4862813877822923e-05,
|
| 376 |
+
"logits/chosen": -1.6853389739990234,
|
| 377 |
+
"logits/rejected": -1.3519870042800903,
|
| 378 |
+
"logps/chosen": -58.672515869140625,
|
| 379 |
+
"logps/rejected": -232.4235382080078,
|
| 380 |
+
"loss": 0.0091,
|
| 381 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 382 |
+
"rewards/chosen": -1.8344595432281494,
|
| 383 |
+
"rewards/margins": 13.702409744262695,
|
| 384 |
+
"rewards/rejected": -15.53686809539795,
|
| 385 |
+
"step": 125
|
| 386 |
+
},
|
| 387 |
+
{
|
| 388 |
+
"epoch": 0.5555555555555556,
|
| 389 |
+
"grad_norm": 0.18359375,
|
| 390 |
+
"learning_rate": 1.4806141921473063e-05,
|
| 391 |
+
"logits/chosen": -1.711216926574707,
|
| 392 |
+
"logits/rejected": -1.331209421157837,
|
| 393 |
+
"logps/chosen": -58.75749969482422,
|
| 394 |
+
"logps/rejected": -241.5647430419922,
|
| 395 |
+
"loss": 0.0058,
|
| 396 |
+
"rewards/accuracies": 1.0,
|
| 397 |
+
"rewards/chosen": -1.8602139949798584,
|
| 398 |
+
"rewards/margins": 14.427963256835938,
|
| 399 |
+
"rewards/rejected": -16.288179397583008,
|
| 400 |
+
"step": 130
|
| 401 |
+
},
|
| 402 |
+
{
|
| 403 |
+
"epoch": 0.5769230769230769,
|
| 404 |
+
"grad_norm": 0.0157470703125,
|
| 405 |
+
"learning_rate": 1.4746583264985202e-05,
|
| 406 |
+
"logits/chosen": -1.73836350440979,
|
| 407 |
+
"logits/rejected": -1.358798623085022,
|
| 408 |
+
"logps/chosen": -57.656578063964844,
|
| 409 |
+
"logps/rejected": -245.814697265625,
|
| 410 |
+
"loss": 0.0054,
|
| 411 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 412 |
+
"rewards/chosen": -1.727595329284668,
|
| 413 |
+
"rewards/margins": 14.943672180175781,
|
| 414 |
+
"rewards/rejected": -16.671268463134766,
|
| 415 |
+
"step": 135
|
| 416 |
+
},
|
| 417 |
+
{
|
| 418 |
+
"epoch": 0.5982905982905983,
|
| 419 |
+
"grad_norm": 0.017578125,
|
| 420 |
+
"learning_rate": 1.468417093868888e-05,
|
| 421 |
+
"logits/chosen": -1.7839100360870361,
|
| 422 |
+
"logits/rejected": -1.4424632787704468,
|
| 423 |
+
"logps/chosen": -55.03651809692383,
|
| 424 |
+
"logps/rejected": -240.14572143554688,
|
| 425 |
+
"loss": 0.0047,
|
| 426 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 427 |
+
"rewards/chosen": -1.5889627933502197,
|
| 428 |
+
"rewards/margins": 14.578252792358398,
|
| 429 |
+
"rewards/rejected": -16.16721534729004,
|
| 430 |
+
"step": 140
|
| 431 |
+
},
|
| 432 |
+
{
|
| 433 |
+
"epoch": 0.6196581196581197,
|
| 434 |
+
"grad_norm": 0.06884765625,
|
| 435 |
+
"learning_rate": 1.4618939555515721e-05,
|
| 436 |
+
"logits/chosen": -1.7428079843521118,
|
| 437 |
+
"logits/rejected": -1.4061057567596436,
|
| 438 |
+
"logps/chosen": -59.45949172973633,
|
| 439 |
+
"logps/rejected": -239.75048828125,
|
| 440 |
+
"loss": 0.0028,
|
| 441 |
+
"rewards/accuracies": 1.0,
|
| 442 |
+
"rewards/chosen": -1.9328705072402954,
|
| 443 |
+
"rewards/margins": 14.28913402557373,
|
| 444 |
+
"rewards/rejected": -16.222003936767578,
|
| 445 |
+
"step": 145
|
| 446 |
+
},
|
| 447 |
+
{
|
| 448 |
+
"epoch": 0.6410256410256411,
|
| 449 |
+
"grad_norm": 0.00933837890625,
|
| 450 |
+
"learning_rate": 1.455092529180363e-05,
|
| 451 |
+
"logits/chosen": -1.7827781438827515,
|
| 452 |
+
"logits/rejected": -1.4337228536605835,
|
| 453 |
+
"logps/chosen": -57.58295440673828,
|
| 454 |
+
"logps/rejected": -243.2423553466797,
|
| 455 |
+
"loss": 0.0003,
|
| 456 |
+
"rewards/accuracies": 1.0,
|
| 457 |
+
"rewards/chosen": -1.8888952732086182,
|
| 458 |
+
"rewards/margins": 14.729533195495605,
|
| 459 |
+
"rewards/rejected": -16.618427276611328,
|
| 460 |
+
"step": 150
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"epoch": 0.6623931623931624,
|
| 464 |
+
"grad_norm": 1.1171875,
|
| 465 |
+
"learning_rate": 1.4480165867233946e-05,
|
| 466 |
+
"logits/chosen": -1.7574710845947266,
|
| 467 |
+
"logits/rejected": -1.4313023090362549,
|
| 468 |
+
"logps/chosen": -61.97917938232422,
|
| 469 |
+
"logps/rejected": -249.78890991210938,
|
| 470 |
+
"loss": 0.0062,
|
| 471 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 472 |
+
"rewards/chosen": -2.226684093475342,
|
| 473 |
+
"rewards/margins": 14.859643936157227,
|
| 474 |
+
"rewards/rejected": -17.086326599121094,
|
| 475 |
+
"step": 155
|
| 476 |
+
},
|
| 477 |
+
{
|
| 478 |
+
"epoch": 0.6837606837606838,
|
| 479 |
+
"grad_norm": 2.59375,
|
| 480 |
+
"learning_rate": 1.440670052391267e-05,
|
| 481 |
+
"logits/chosen": -1.776049017906189,
|
| 482 |
+
"logits/rejected": -1.4136337041854858,
|
| 483 |
+
"logps/chosen": -58.725502014160156,
|
| 484 |
+
"logps/rejected": -239.2755126953125,
|
| 485 |
+
"loss": 0.0046,
|
| 486 |
+
"rewards/accuracies": 1.0,
|
| 487 |
+
"rewards/chosen": -2.0192863941192627,
|
| 488 |
+
"rewards/margins": 14.30653190612793,
|
| 489 |
+
"rewards/rejected": -16.32581901550293,
|
| 490 |
+
"step": 160
|
| 491 |
+
},
|
| 492 |
+
{
|
| 493 |
+
"epoch": 0.7051282051282052,
|
| 494 |
+
"grad_norm": 1.7734375,
|
| 495 |
+
"learning_rate": 1.4330570004607398e-05,
|
| 496 |
+
"logits/chosen": -1.8287827968597412,
|
| 497 |
+
"logits/rejected": -1.4543850421905518,
|
| 498 |
+
"logps/chosen": -57.99534225463867,
|
| 499 |
+
"logps/rejected": -249.4701690673828,
|
| 500 |
+
"loss": 0.0035,
|
| 501 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 502 |
+
"rewards/chosen": -1.8924716711044312,
|
| 503 |
+
"rewards/margins": 15.258901596069336,
|
| 504 |
+
"rewards/rejected": -17.15137481689453,
|
| 505 |
+
"step": 165
|
| 506 |
+
},
|
| 507 |
+
{
|
| 508 |
+
"epoch": 0.7264957264957265,
|
| 509 |
+
"grad_norm": 1.359375,
|
| 510 |
+
"learning_rate": 1.4251816530151986e-05,
|
| 511 |
+
"logits/chosen": -1.7740707397460938,
|
| 512 |
+
"logits/rejected": -1.4325586557388306,
|
| 513 |
+
"logps/chosen": -60.86432647705078,
|
| 514 |
+
"logps/rejected": -249.397216796875,
|
| 515 |
+
"loss": 0.0032,
|
| 516 |
+
"rewards/accuracies": 1.0,
|
| 517 |
+
"rewards/chosen": -2.059633255004883,
|
| 518 |
+
"rewards/margins": 15.068583488464355,
|
| 519 |
+
"rewards/rejected": -17.128215789794922,
|
| 520 |
+
"step": 170
|
| 521 |
+
},
|
| 522 |
+
{
|
| 523 |
+
"epoch": 0.7478632478632479,
|
| 524 |
+
"grad_norm": 0.060546875,
|
| 525 |
+
"learning_rate": 1.4170483776031526e-05,
|
| 526 |
+
"logits/chosen": -1.7101682424545288,
|
| 527 |
+
"logits/rejected": -1.3118056058883667,
|
| 528 |
+
"logps/chosen": -60.89410400390625,
|
| 529 |
+
"logps/rejected": -252.2559814453125,
|
| 530 |
+
"loss": 0.0002,
|
| 531 |
+
"rewards/accuracies": 1.0,
|
| 532 |
+
"rewards/chosen": -2.1261119842529297,
|
| 533 |
+
"rewards/margins": 15.250958442687988,
|
| 534 |
+
"rewards/rejected": -17.377071380615234,
|
| 535 |
+
"step": 175
|
| 536 |
+
},
|
| 537 |
+
{
|
| 538 |
+
"epoch": 0.7692307692307693,
|
| 539 |
+
"grad_norm": 0.031494140625,
|
| 540 |
+
"learning_rate": 1.4086616848160574e-05,
|
| 541 |
+
"logits/chosen": -1.6894880533218384,
|
| 542 |
+
"logits/rejected": -1.2855875492095947,
|
| 543 |
+
"logps/chosen": -66.87565612792969,
|
| 544 |
+
"logps/rejected": -245.99557495117188,
|
| 545 |
+
"loss": 0.0003,
|
| 546 |
+
"rewards/accuracies": 1.0,
|
| 547 |
+
"rewards/chosen": -2.7102391719818115,
|
| 548 |
+
"rewards/margins": 14.112167358398438,
|
| 549 |
+
"rewards/rejected": -16.822406768798828,
|
| 550 |
+
"step": 180
|
| 551 |
+
},
|
| 552 |
+
{
|
| 553 |
+
"epoch": 0.7905982905982906,
|
| 554 |
+
"grad_norm": 0.09912109375,
|
| 555 |
+
"learning_rate": 1.4000262257868096e-05,
|
| 556 |
+
"logits/chosen": -1.6548315286636353,
|
| 557 |
+
"logits/rejected": -1.192333459854126,
|
| 558 |
+
"logps/chosen": -64.77526092529297,
|
| 559 |
+
"logps/rejected": -249.92196655273438,
|
| 560 |
+
"loss": 0.0027,
|
| 561 |
+
"rewards/accuracies": 1.0,
|
| 562 |
+
"rewards/chosen": -2.546926975250244,
|
| 563 |
+
"rewards/margins": 14.788670539855957,
|
| 564 |
+
"rewards/rejected": -17.33559799194336,
|
| 565 |
+
"step": 185
|
| 566 |
+
},
|
| 567 |
+
{
|
| 568 |
+
"epoch": 0.811965811965812,
|
| 569 |
+
"grad_norm": 1.1953125,
|
| 570 |
+
"learning_rate": 1.3911467896102994e-05,
|
| 571 |
+
"logits/chosen": -1.6022329330444336,
|
| 572 |
+
"logits/rejected": -1.1850754022598267,
|
| 573 |
+
"logps/chosen": -60.396514892578125,
|
| 574 |
+
"logps/rejected": -247.5742645263672,
|
| 575 |
+
"loss": 0.0054,
|
| 576 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 577 |
+
"rewards/chosen": -2.1046481132507324,
|
| 578 |
+
"rewards/margins": 15.000350952148438,
|
| 579 |
+
"rewards/rejected": -17.104999542236328,
|
| 580 |
+
"step": 190
|
| 581 |
+
},
|
| 582 |
+
{
|
| 583 |
+
"epoch": 0.8333333333333334,
|
| 584 |
+
"grad_norm": 0.10400390625,
|
| 585 |
+
"learning_rate": 1.3820283006874503e-05,
|
| 586 |
+
"logits/chosen": -1.563820719718933,
|
| 587 |
+
"logits/rejected": -1.137289047241211,
|
| 588 |
+
"logps/chosen": -65.57173156738281,
|
| 589 |
+
"logps/rejected": -252.38955688476562,
|
| 590 |
+
"loss": 0.0085,
|
| 591 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 592 |
+
"rewards/chosen": -2.3488571643829346,
|
| 593 |
+
"rewards/margins": 15.110745429992676,
|
| 594 |
+
"rewards/rejected": -17.45960235595703,
|
| 595 |
+
"step": 195
|
| 596 |
+
},
|
| 597 |
+
{
|
| 598 |
+
"epoch": 0.8547008547008547,
|
| 599 |
+
"grad_norm": 0.1279296875,
|
| 600 |
+
"learning_rate": 1.372675815994221e-05,
|
| 601 |
+
"logits/chosen": -1.4945417642593384,
|
| 602 |
+
"logits/rejected": -1.038334846496582,
|
| 603 |
+
"logps/chosen": -54.2162971496582,
|
| 604 |
+
"logps/rejected": -254.0222625732422,
|
| 605 |
+
"loss": 0.0003,
|
| 606 |
+
"rewards/accuracies": 1.0,
|
| 607 |
+
"rewards/chosen": -1.6657600402832031,
|
| 608 |
+
"rewards/margins": 16.079242706298828,
|
| 609 |
+
"rewards/rejected": -17.745004653930664,
|
| 610 |
+
"step": 200
|
| 611 |
+
},
|
| 612 |
+
{
|
| 613 |
+
"epoch": 0.8760683760683761,
|
| 614 |
+
"grad_norm": 14.625,
|
| 615 |
+
"learning_rate": 1.3630945222770829e-05,
|
| 616 |
+
"logits/chosen": -1.5167819261550903,
|
| 617 |
+
"logits/rejected": -1.0430529117584229,
|
| 618 |
+
"logps/chosen": -60.5411262512207,
|
| 619 |
+
"logps/rejected": -259.63555908203125,
|
| 620 |
+
"loss": 0.0192,
|
| 621 |
+
"rewards/accuracies": 0.9916666746139526,
|
| 622 |
+
"rewards/chosen": -2.0362796783447266,
|
| 623 |
+
"rewards/margins": 16.097753524780273,
|
| 624 |
+
"rewards/rejected": -18.134033203125,
|
| 625 |
+
"step": 205
|
| 626 |
+
},
|
| 627 |
+
{
|
| 628 |
+
"epoch": 0.8974358974358975,
|
| 629 |
+
"grad_norm": 0.2119140625,
|
| 630 |
+
"learning_rate": 1.3532897331765301e-05,
|
| 631 |
+
"logits/chosen": -1.5572597980499268,
|
| 632 |
+
"logits/rejected": -1.053264856338501,
|
| 633 |
+
"logps/chosen": -59.34284591674805,
|
| 634 |
+
"logps/rejected": -260.30206298828125,
|
| 635 |
+
"loss": 0.0151,
|
| 636 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 637 |
+
"rewards/chosen": -1.9360477924346924,
|
| 638 |
+
"rewards/margins": 16.303897857666016,
|
| 639 |
+
"rewards/rejected": -18.239948272705078,
|
| 640 |
+
"step": 210
|
| 641 |
+
},
|
| 642 |
+
{
|
| 643 |
+
"epoch": 0.9188034188034188,
|
| 644 |
+
"grad_norm": 0.0849609375,
|
| 645 |
+
"learning_rate": 1.3432668862802134e-05,
|
| 646 |
+
"logits/chosen": -1.4950945377349854,
|
| 647 |
+
"logits/rejected": -1.014696478843689,
|
| 648 |
+
"logps/chosen": -57.9535026550293,
|
| 649 |
+
"logps/rejected": -256.272705078125,
|
| 650 |
+
"loss": 0.0001,
|
| 651 |
+
"rewards/accuracies": 1.0,
|
| 652 |
+
"rewards/chosen": -1.7511317729949951,
|
| 653 |
+
"rewards/margins": 15.973749160766602,
|
| 654 |
+
"rewards/rejected": -17.724878311157227,
|
| 655 |
+
"step": 215
|
| 656 |
+
},
|
| 657 |
+
{
|
| 658 |
+
"epoch": 0.9401709401709402,
|
| 659 |
+
"grad_norm": 0.08642578125,
|
| 660 |
+
"learning_rate": 1.3330315401073371e-05,
|
| 661 |
+
"logits/chosen": -1.5073899030685425,
|
| 662 |
+
"logits/rejected": -1.0442817211151123,
|
| 663 |
+
"logps/chosen": -59.66225051879883,
|
| 664 |
+
"logps/rejected": -249.38534545898438,
|
| 665 |
+
"loss": 0.002,
|
| 666 |
+
"rewards/accuracies": 1.0,
|
| 667 |
+
"rewards/chosen": -1.872859001159668,
|
| 668 |
+
"rewards/margins": 15.302103042602539,
|
| 669 |
+
"rewards/rejected": -17.17496109008789,
|
| 670 |
+
"step": 220
|
| 671 |
+
},
|
| 672 |
+
{
|
| 673 |
+
"epoch": 0.9615384615384616,
|
| 674 |
+
"grad_norm": 2.625,
|
| 675 |
+
"learning_rate": 1.3225893710259887e-05,
|
| 676 |
+
"logits/chosen": -1.3290693759918213,
|
| 677 |
+
"logits/rejected": -0.8046108484268188,
|
| 678 |
+
"logps/chosen": -60.98704147338867,
|
| 679 |
+
"logps/rejected": -252.48495483398438,
|
| 680 |
+
"loss": 0.0065,
|
| 681 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 682 |
+
"rewards/chosen": -2.179378032684326,
|
| 683 |
+
"rewards/margins": 15.35717487335205,
|
| 684 |
+
"rewards/rejected": -17.53655433654785,
|
| 685 |
+
"step": 225
|
| 686 |
+
},
|
| 687 |
+
{
|
| 688 |
+
"epoch": 0.9829059829059829,
|
| 689 |
+
"grad_norm": 1.359375,
|
| 690 |
+
"learning_rate": 1.3119461701051105e-05,
|
| 691 |
+
"logits/chosen": -1.4031749963760376,
|
| 692 |
+
"logits/rejected": -0.8651553988456726,
|
| 693 |
+
"logps/chosen": -57.99671173095703,
|
| 694 |
+
"logps/rejected": -251.2525634765625,
|
| 695 |
+
"loss": 0.0006,
|
| 696 |
+
"rewards/accuracies": 1.0,
|
| 697 |
+
"rewards/chosen": -2.0349326133728027,
|
| 698 |
+
"rewards/margins": 15.621841430664062,
|
| 699 |
+
"rewards/rejected": -17.656774520874023,
|
| 700 |
+
"step": 230
|
| 701 |
+
},
|
| 702 |
+
{
|
| 703 |
+
"epoch": 0.9957264957264957,
|
| 704 |
+
"eval_logits/chosen": -1.4502625465393066,
|
| 705 |
+
"eval_logits/rejected": -0.9043333530426025,
|
| 706 |
+
"eval_logps/chosen": -61.05683517456055,
|
| 707 |
+
"eval_logps/rejected": -259.54730224609375,
|
| 708 |
+
"eval_loss": 0.0003871396475005895,
|
| 709 |
+
"eval_rewards/accuracies": 1.0,
|
| 710 |
+
"eval_rewards/chosen": -2.1613101959228516,
|
| 711 |
+
"eval_rewards/margins": 15.991730690002441,
|
| 712 |
+
"eval_rewards/rejected": -18.15304183959961,
|
| 713 |
+
"eval_runtime": 9.8475,
|
| 714 |
+
"eval_samples_per_second": 20.31,
|
| 715 |
+
"eval_steps_per_second": 20.31,
|
| 716 |
+
"step": 233
|
| 717 |
+
},
|
| 718 |
+
{
|
| 719 |
+
"epoch": 1.0042735042735043,
|
| 720 |
+
"grad_norm": 0.0341796875,
|
| 721 |
+
"learning_rate": 1.3011078399028605e-05,
|
| 722 |
+
"logits/chosen": -1.430901288986206,
|
| 723 |
+
"logits/rejected": -0.9248638153076172,
|
| 724 |
+
"logps/chosen": -60.4477424621582,
|
| 725 |
+
"logps/rejected": -265.5071716308594,
|
| 726 |
+
"loss": 0.0036,
|
| 727 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 728 |
+
"rewards/chosen": -2.0871260166168213,
|
| 729 |
+
"rewards/margins": 16.610048294067383,
|
| 730 |
+
"rewards/rejected": -18.697174072265625,
|
| 731 |
+
"step": 235
|
| 732 |
+
},
|
| 733 |
+
{
|
| 734 |
+
"epoch": 1.0256410256410255,
|
| 735 |
+
"grad_norm": 0.058349609375,
|
| 736 |
+
"learning_rate": 1.2900803911931431e-05,
|
| 737 |
+
"logits/chosen": -1.4457504749298096,
|
| 738 |
+
"logits/rejected": -0.9034906625747681,
|
| 739 |
+
"logps/chosen": -59.68601608276367,
|
| 740 |
+
"logps/rejected": -263.2565612792969,
|
| 741 |
+
"loss": 0.0001,
|
| 742 |
+
"rewards/accuracies": 1.0,
|
| 743 |
+
"rewards/chosen": -1.9931907653808594,
|
| 744 |
+
"rewards/margins": 16.479150772094727,
|
| 745 |
+
"rewards/rejected": -18.472341537475586,
|
| 746 |
+
"step": 240
|
| 747 |
+
},
|
| 748 |
+
{
|
| 749 |
+
"epoch": 1.047008547008547,
|
| 750 |
+
"grad_norm": 0.037841796875,
|
| 751 |
+
"learning_rate": 1.2788699396321252e-05,
|
| 752 |
+
"logits/chosen": -1.39047110080719,
|
| 753 |
+
"logits/rejected": -0.8637332916259766,
|
| 754 |
+
"logps/chosen": -57.831581115722656,
|
| 755 |
+
"logps/rejected": -255.6790771484375,
|
| 756 |
+
"loss": 0.0001,
|
| 757 |
+
"rewards/accuracies": 1.0,
|
| 758 |
+
"rewards/chosen": -1.925370454788208,
|
| 759 |
+
"rewards/margins": 16.076345443725586,
|
| 760 |
+
"rewards/rejected": -18.00171661376953,
|
| 761 |
+
"step": 245
|
| 762 |
+
},
|
| 763 |
+
{
|
| 764 |
+
"epoch": 1.0683760683760684,
|
| 765 |
+
"grad_norm": 0.0002689361572265625,
|
| 766 |
+
"learning_rate": 1.2674827023665853e-05,
|
| 767 |
+
"logits/chosen": -1.4293615818023682,
|
| 768 |
+
"logits/rejected": -0.9234841465950012,
|
| 769 |
+
"logps/chosen": -61.868263244628906,
|
| 770 |
+
"logps/rejected": -270.99920654296875,
|
| 771 |
+
"loss": 0.0008,
|
| 772 |
+
"rewards/accuracies": 1.0,
|
| 773 |
+
"rewards/chosen": -2.230020046234131,
|
| 774 |
+
"rewards/margins": 17.106487274169922,
|
| 775 |
+
"rewards/rejected": -19.336505889892578,
|
| 776 |
+
"step": 250
|
| 777 |
+
},
|
| 778 |
+
{
|
| 779 |
+
"epoch": 1.0897435897435896,
|
| 780 |
+
"grad_norm": 0.08154296875,
|
| 781 |
+
"learning_rate": 1.255924994585978e-05,
|
| 782 |
+
"logits/chosen": -1.4079844951629639,
|
| 783 |
+
"logits/rejected": -0.8683494329452515,
|
| 784 |
+
"logps/chosen": -63.99333953857422,
|
| 785 |
+
"logps/rejected": -266.74462890625,
|
| 786 |
+
"loss": 0.0001,
|
| 787 |
+
"rewards/accuracies": 1.0,
|
| 788 |
+
"rewards/chosen": -2.376952886581421,
|
| 789 |
+
"rewards/margins": 16.51467514038086,
|
| 790 |
+
"rewards/rejected": -18.89162826538086,
|
| 791 |
+
"step": 255
|
| 792 |
+
},
|
| 793 |
+
{
|
| 794 |
+
"epoch": 1.1111111111111112,
|
| 795 |
+
"grad_norm": 0.00147247314453125,
|
| 796 |
+
"learning_rate": 1.2442032260201255e-05,
|
| 797 |
+
"logits/chosen": -1.400887131690979,
|
| 798 |
+
"logits/rejected": -0.890865683555603,
|
| 799 |
+
"logps/chosen": -62.627723693847656,
|
| 800 |
+
"logps/rejected": -263.9581604003906,
|
| 801 |
+
"loss": 0.0001,
|
| 802 |
+
"rewards/accuracies": 1.0,
|
| 803 |
+
"rewards/chosen": -2.308192729949951,
|
| 804 |
+
"rewards/margins": 16.26297950744629,
|
| 805 |
+
"rewards/rejected": -18.571170806884766,
|
| 806 |
+
"step": 260
|
| 807 |
+
},
|
| 808 |
+
{
|
| 809 |
+
"epoch": 1.1324786324786325,
|
| 810 |
+
"grad_norm": 0.1982421875,
|
| 811 |
+
"learning_rate": 1.2323238973844796e-05,
|
| 812 |
+
"logits/chosen": -1.438955545425415,
|
| 813 |
+
"logits/rejected": -0.9066799283027649,
|
| 814 |
+
"logps/chosen": -63.32421875,
|
| 815 |
+
"logps/rejected": -273.82049560546875,
|
| 816 |
+
"loss": 0.0001,
|
| 817 |
+
"rewards/accuracies": 1.0,
|
| 818 |
+
"rewards/chosen": -2.295706272125244,
|
| 819 |
+
"rewards/margins": 17.04709243774414,
|
| 820 |
+
"rewards/rejected": -19.342798233032227,
|
| 821 |
+
"step": 265
|
| 822 |
+
},
|
| 823 |
+
{
|
| 824 |
+
"epoch": 1.1538461538461537,
|
| 825 |
+
"grad_norm": 0.011962890625,
|
| 826 |
+
"learning_rate": 1.2202935967749212e-05,
|
| 827 |
+
"logits/chosen": -1.3738555908203125,
|
| 828 |
+
"logits/rejected": -0.8493305444717407,
|
| 829 |
+
"logps/chosen": -67.02043151855469,
|
| 830 |
+
"logps/rejected": -268.3296813964844,
|
| 831 |
+
"loss": 0.0003,
|
| 832 |
+
"rewards/accuracies": 1.0,
|
| 833 |
+
"rewards/chosen": -2.584259510040283,
|
| 834 |
+
"rewards/margins": 16.276723861694336,
|
| 835 |
+
"rewards/rejected": -18.86098289489746,
|
| 836 |
+
"step": 270
|
| 837 |
+
},
|
| 838 |
+
{
|
| 839 |
+
"epoch": 1.1752136752136753,
|
| 840 |
+
"grad_norm": 1.765625,
|
| 841 |
+
"learning_rate": 1.2081189960141038e-05,
|
| 842 |
+
"logits/chosen": -1.424109697341919,
|
| 843 |
+
"logits/rejected": -0.8860370516777039,
|
| 844 |
+
"logps/chosen": -61.854644775390625,
|
| 845 |
+
"logps/rejected": -267.197021484375,
|
| 846 |
+
"loss": 0.0013,
|
| 847 |
+
"rewards/accuracies": 1.0,
|
| 848 |
+
"rewards/chosen": -2.1506810188293457,
|
| 849 |
+
"rewards/margins": 16.72945213317871,
|
| 850 |
+
"rewards/rejected": -18.880136489868164,
|
| 851 |
+
"step": 275
|
| 852 |
+
},
|
| 853 |
+
{
|
| 854 |
+
"epoch": 1.1965811965811965,
|
| 855 |
+
"grad_norm": 0.185546875,
|
| 856 |
+
"learning_rate": 1.1958068469513604e-05,
|
| 857 |
+
"logits/chosen": -1.4285290241241455,
|
| 858 |
+
"logits/rejected": -0.9055356979370117,
|
| 859 |
+
"logps/chosen": -65.09664916992188,
|
| 860 |
+
"logps/rejected": -277.3038635253906,
|
| 861 |
+
"loss": 0.0015,
|
| 862 |
+
"rewards/accuracies": 1.0,
|
| 863 |
+
"rewards/chosen": -2.459376096725464,
|
| 864 |
+
"rewards/margins": 17.357646942138672,
|
| 865 |
+
"rewards/rejected": -19.817026138305664,
|
| 866 |
+
"step": 280
|
| 867 |
+
},
|
| 868 |
+
{
|
| 869 |
+
"epoch": 1.217948717948718,
|
| 870 |
+
"grad_norm": 0.0059814453125,
|
| 871 |
+
"learning_rate": 1.1833639777182316e-05,
|
| 872 |
+
"logits/chosen": -1.306983232498169,
|
| 873 |
+
"logits/rejected": -0.7688174247741699,
|
| 874 |
+
"logps/chosen": -59.734840393066406,
|
| 875 |
+
"logps/rejected": -272.2923889160156,
|
| 876 |
+
"loss": 0.0,
|
| 877 |
+
"rewards/accuracies": 1.0,
|
| 878 |
+
"rewards/chosen": -2.1238536834716797,
|
| 879 |
+
"rewards/margins": 17.397279739379883,
|
| 880 |
+
"rewards/rejected": -19.52113151550293,
|
| 881 |
+
"step": 285
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"epoch": 1.2393162393162394,
|
| 885 |
+
"grad_norm": 0.004547119140625,
|
| 886 |
+
"learning_rate": 1.170797288941685e-05,
|
| 887 |
+
"logits/chosen": -1.2420094013214111,
|
| 888 |
+
"logits/rejected": -0.7252348065376282,
|
| 889 |
+
"logps/chosen": -66.8678207397461,
|
| 890 |
+
"logps/rejected": -281.2005920410156,
|
| 891 |
+
"loss": 0.0007,
|
| 892 |
+
"rewards/accuracies": 1.0,
|
| 893 |
+
"rewards/chosen": -2.6819348335266113,
|
| 894 |
+
"rewards/margins": 17.403133392333984,
|
| 895 |
+
"rewards/rejected": -20.085067749023438,
|
| 896 |
+
"step": 290
|
| 897 |
+
},
|
| 898 |
+
{
|
| 899 |
+
"epoch": 1.2606837606837606,
|
| 900 |
+
"grad_norm": 0.8828125,
|
| 901 |
+
"learning_rate": 1.1581137499171342e-05,
|
| 902 |
+
"logits/chosen": -1.2711966037750244,
|
| 903 |
+
"logits/rejected": -0.7393882870674133,
|
| 904 |
+
"logps/chosen": -67.3682861328125,
|
| 905 |
+
"logps/rejected": -272.0533142089844,
|
| 906 |
+
"loss": 0.0012,
|
| 907 |
+
"rewards/accuracies": 1.0,
|
| 908 |
+
"rewards/chosen": -2.7543585300445557,
|
| 909 |
+
"rewards/margins": 16.66292953491211,
|
| 910 |
+
"rewards/rejected": -19.41728973388672,
|
| 911 |
+
"step": 295
|
| 912 |
+
},
|
| 913 |
+
{
|
| 914 |
+
"epoch": 1.282051282051282,
|
| 915 |
+
"grad_norm": 0.01080322265625,
|
| 916 |
+
"learning_rate": 1.145320394743371e-05,
|
| 917 |
+
"logits/chosen": -1.2599390745162964,
|
| 918 |
+
"logits/rejected": -0.7466105222702026,
|
| 919 |
+
"logps/chosen": -65.01288604736328,
|
| 920 |
+
"logps/rejected": -266.981689453125,
|
| 921 |
+
"loss": 0.0002,
|
| 922 |
+
"rewards/accuracies": 1.0,
|
| 923 |
+
"rewards/chosen": -2.5523085594177246,
|
| 924 |
+
"rewards/margins": 16.560453414916992,
|
| 925 |
+
"rewards/rejected": -19.112764358520508,
|
| 926 |
+
"step": 300
|
| 927 |
+
},
|
| 928 |
+
{
|
| 929 |
+
"epoch": 1.3034188034188035,
|
| 930 |
+
"grad_norm": 0.005706787109375,
|
| 931 |
+
"learning_rate": 1.1324243184215622e-05,
|
| 932 |
+
"logits/chosen": -1.268808364868164,
|
| 933 |
+
"logits/rejected": -0.7630107998847961,
|
| 934 |
+
"logps/chosen": -67.40531921386719,
|
| 935 |
+
"logps/rejected": -282.67205810546875,
|
| 936 |
+
"loss": 0.0001,
|
| 937 |
+
"rewards/accuracies": 1.0,
|
| 938 |
+
"rewards/chosen": -2.892889976501465,
|
| 939 |
+
"rewards/margins": 17.570592880249023,
|
| 940 |
+
"rewards/rejected": -20.463483810424805,
|
| 941 |
+
"step": 305
|
| 942 |
+
},
|
| 943 |
+
{
|
| 944 |
+
"epoch": 1.3247863247863247,
|
| 945 |
+
"grad_norm": 0.00238037109375,
|
| 946 |
+
"learning_rate": 1.1194326729204686e-05,
|
| 947 |
+
"logits/chosen": -1.2974001169204712,
|
| 948 |
+
"logits/rejected": -0.7796735167503357,
|
| 949 |
+
"logps/chosen": -67.35159301757812,
|
| 950 |
+
"logps/rejected": -279.50341796875,
|
| 951 |
+
"loss": 0.0,
|
| 952 |
+
"rewards/accuracies": 1.0,
|
| 953 |
+
"rewards/chosen": -2.8483831882476807,
|
| 954 |
+
"rewards/margins": 17.4374942779541,
|
| 955 |
+
"rewards/rejected": -20.285879135131836,
|
| 956 |
+
"step": 310
|
| 957 |
+
},
|
| 958 |
+
{
|
| 959 |
+
"epoch": 1.3461538461538463,
|
| 960 |
+
"grad_norm": 0.021484375,
|
| 961 |
+
"learning_rate": 1.1063526632100717e-05,
|
| 962 |
+
"logits/chosen": -1.2823902368545532,
|
| 963 |
+
"logits/rejected": -0.8006687164306641,
|
| 964 |
+
"logps/chosen": -71.19620513916016,
|
| 965 |
+
"logps/rejected": -273.39569091796875,
|
| 966 |
+
"loss": 0.0031,
|
| 967 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 968 |
+
"rewards/chosen": -3.098231077194214,
|
| 969 |
+
"rewards/margins": 16.550390243530273,
|
| 970 |
+
"rewards/rejected": -19.648624420166016,
|
| 971 |
+
"step": 315
|
| 972 |
+
},
|
| 973 |
+
{
|
| 974 |
+
"epoch": 1.3675213675213675,
|
| 975 |
+
"grad_norm": 0.022216796875,
|
| 976 |
+
"learning_rate": 1.0931915432658055e-05,
|
| 977 |
+
"logits/chosen": -1.2740647792816162,
|
| 978 |
+
"logits/rejected": -0.7717633843421936,
|
| 979 |
+
"logps/chosen": -68.88563537597656,
|
| 980 |
+
"logps/rejected": -277.2671813964844,
|
| 981 |
+
"loss": 0.0001,
|
| 982 |
+
"rewards/accuracies": 1.0,
|
| 983 |
+
"rewards/chosen": -2.923701763153076,
|
| 984 |
+
"rewards/margins": 17.007097244262695,
|
| 985 |
+
"rewards/rejected": -19.930797576904297,
|
| 986 |
+
"step": 320
|
| 987 |
+
},
|
| 988 |
+
{
|
| 989 |
+
"epoch": 1.3888888888888888,
|
| 990 |
+
"grad_norm": 0.048095703125,
|
| 991 |
+
"learning_rate": 1.0799566120456133e-05,
|
| 992 |
+
"logits/chosen": -1.2586907148361206,
|
| 993 |
+
"logits/rejected": -0.7510126829147339,
|
| 994 |
+
"logps/chosen": -71.87751770019531,
|
| 995 |
+
"logps/rejected": -275.9109191894531,
|
| 996 |
+
"loss": 0.0001,
|
| 997 |
+
"rewards/accuracies": 1.0,
|
| 998 |
+
"rewards/chosen": -3.1944289207458496,
|
| 999 |
+
"rewards/margins": 16.62671661376953,
|
| 1000 |
+
"rewards/rejected": -19.82114601135254,
|
| 1001 |
+
"step": 325
|
| 1002 |
+
},
|
| 1003 |
+
{
|
| 1004 |
+
"epoch": 1.4102564102564101,
|
| 1005 |
+
"grad_norm": 0.0052490234375,
|
| 1006 |
+
"learning_rate": 1.066655209442054e-05,
|
| 1007 |
+
"logits/chosen": -1.280989646911621,
|
| 1008 |
+
"logits/rejected": -0.772638201713562,
|
| 1009 |
+
"logps/chosen": -71.73796081542969,
|
| 1010 |
+
"logps/rejected": -279.8167419433594,
|
| 1011 |
+
"loss": 0.0021,
|
| 1012 |
+
"rewards/accuracies": 1.0,
|
| 1013 |
+
"rewards/chosen": -3.172919511795044,
|
| 1014 |
+
"rewards/margins": 16.89200210571289,
|
| 1015 |
+
"rewards/rejected": -20.064922332763672,
|
| 1016 |
+
"step": 330
|
| 1017 |
+
},
|
| 1018 |
+
{
|
| 1019 |
+
"epoch": 1.4316239316239316,
|
| 1020 |
+
"grad_norm": 0.007720947265625,
|
| 1021 |
+
"learning_rate": 1.0532947122117101e-05,
|
| 1022 |
+
"logits/chosen": -1.27340567111969,
|
| 1023 |
+
"logits/rejected": -0.7604951858520508,
|
| 1024 |
+
"logps/chosen": -69.45366668701172,
|
| 1025 |
+
"logps/rejected": -278.88409423828125,
|
| 1026 |
+
"loss": 0.0,
|
| 1027 |
+
"rewards/accuracies": 1.0,
|
| 1028 |
+
"rewards/chosen": -2.9620535373687744,
|
| 1029 |
+
"rewards/margins": 17.169063568115234,
|
| 1030 |
+
"rewards/rejected": -20.13111686706543,
|
| 1031 |
+
"step": 335
|
| 1032 |
+
},
|
| 1033 |
+
{
|
| 1034 |
+
"epoch": 1.452991452991453,
|
| 1035 |
+
"grad_norm": 0.046630859375,
|
| 1036 |
+
"learning_rate": 1.0398825298841499e-05,
|
| 1037 |
+
"logits/chosen": -1.2809860706329346,
|
| 1038 |
+
"logits/rejected": -0.838448166847229,
|
| 1039 |
+
"logps/chosen": -72.61773681640625,
|
| 1040 |
+
"logps/rejected": -281.2303466796875,
|
| 1041 |
+
"loss": 0.0011,
|
| 1042 |
+
"rewards/accuracies": 1.0,
|
| 1043 |
+
"rewards/chosen": -3.1478874683380127,
|
| 1044 |
+
"rewards/margins": 17.092044830322266,
|
| 1045 |
+
"rewards/rejected": -20.239933013916016,
|
| 1046 |
+
"step": 340
|
| 1047 |
+
},
|
| 1048 |
+
{
|
| 1049 |
+
"epoch": 1.4743589743589745,
|
| 1050 |
+
"grad_norm": 0.01025390625,
|
| 1051 |
+
"learning_rate": 1.0264261006527144e-05,
|
| 1052 |
+
"logits/chosen": -1.301695704460144,
|
| 1053 |
+
"logits/rejected": -0.8668543100357056,
|
| 1054 |
+
"logps/chosen": -69.15229797363281,
|
| 1055 |
+
"logps/rejected": -270.77130126953125,
|
| 1056 |
+
"loss": 0.0035,
|
| 1057 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1058 |
+
"rewards/chosen": -2.9070522785186768,
|
| 1059 |
+
"rewards/margins": 16.549983978271484,
|
| 1060 |
+
"rewards/rejected": -19.457035064697266,
|
| 1061 |
+
"step": 345
|
| 1062 |
+
},
|
| 1063 |
+
{
|
| 1064 |
+
"epoch": 1.4957264957264957,
|
| 1065 |
+
"grad_norm": 0.5625,
|
| 1066 |
+
"learning_rate": 1.0129328872494075e-05,
|
| 1067 |
+
"logits/chosen": -1.385507345199585,
|
| 1068 |
+
"logits/rejected": -0.9304911494255066,
|
| 1069 |
+
"logps/chosen": -70.08064270019531,
|
| 1070 |
+
"logps/rejected": -279.90045166015625,
|
| 1071 |
+
"loss": 0.003,
|
| 1072 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1073 |
+
"rewards/chosen": -3.0408215522766113,
|
| 1074 |
+
"rewards/margins": 17.223588943481445,
|
| 1075 |
+
"rewards/rejected": -20.2644100189209,
|
| 1076 |
+
"step": 350
|
| 1077 |
+
},
|
| 1078 |
+
{
|
| 1079 |
+
"epoch": 1.517094017094017,
|
| 1080 |
+
"grad_norm": 0.12890625,
|
| 1081 |
+
"learning_rate": 9.994103728061786e-06,
|
| 1082 |
+
"logits/chosen": -1.3539297580718994,
|
| 1083 |
+
"logits/rejected": -0.8995206952095032,
|
| 1084 |
+
"logps/chosen": -71.8431625366211,
|
| 1085 |
+
"logps/rejected": -278.0020446777344,
|
| 1086 |
+
"loss": 0.0003,
|
| 1087 |
+
"rewards/accuracies": 1.0,
|
| 1088 |
+
"rewards/chosen": -3.1649577617645264,
|
| 1089 |
+
"rewards/margins": 16.83130645751953,
|
| 1090 |
+
"rewards/rejected": -19.99626350402832,
|
| 1091 |
+
"step": 355
|
| 1092 |
+
},
|
| 1093 |
+
{
|
| 1094 |
+
"epoch": 1.5384615384615383,
|
| 1095 |
+
"grad_norm": 0.006134033203125,
|
| 1096 |
+
"learning_rate": 9.858660567048902e-06,
|
| 1097 |
+
"logits/chosen": -1.3628873825073242,
|
| 1098 |
+
"logits/rejected": -0.8617550730705261,
|
| 1099 |
+
"logps/chosen": -73.44374084472656,
|
| 1100 |
+
"logps/rejected": -291.49237060546875,
|
| 1101 |
+
"loss": 0.0,
|
| 1102 |
+
"rewards/accuracies": 1.0,
|
| 1103 |
+
"rewards/chosen": -3.3345115184783936,
|
| 1104 |
+
"rewards/margins": 17.942880630493164,
|
| 1105 |
+
"rewards/rejected": -21.277393341064453,
|
| 1106 |
+
"step": 360
|
| 1107 |
+
},
|
| 1108 |
+
{
|
| 1109 |
+
"epoch": 1.5598290598290598,
|
| 1110 |
+
"grad_norm": 0.036376953125,
|
| 1111 |
+
"learning_rate": 9.72307450418274e-06,
|
| 1112 |
+
"logits/chosen": -1.3687984943389893,
|
| 1113 |
+
"logits/rejected": -0.8952552676200867,
|
| 1114 |
+
"logps/chosen": -68.8270492553711,
|
| 1115 |
+
"logps/rejected": -276.6328125,
|
| 1116 |
+
"loss": 0.0,
|
| 1117 |
+
"rewards/accuracies": 1.0,
|
| 1118 |
+
"rewards/chosen": -2.877126932144165,
|
| 1119 |
+
"rewards/margins": 17.08051872253418,
|
| 1120 |
+
"rewards/rejected": -19.957645416259766,
|
| 1121 |
+
"step": 365
|
| 1122 |
+
},
|
| 1123 |
+
{
|
| 1124 |
+
"epoch": 1.5811965811965814,
|
| 1125 |
+
"grad_norm": 0.400390625,
|
| 1126 |
+
"learning_rate": 9.587420733441835e-06,
|
| 1127 |
+
"logits/chosen": -1.3641754388809204,
|
| 1128 |
+
"logits/rejected": -0.9082571864128113,
|
| 1129 |
+
"logps/chosen": -67.42304992675781,
|
| 1130 |
+
"logps/rejected": -281.8648986816406,
|
| 1131 |
+
"loss": 0.0001,
|
| 1132 |
+
"rewards/accuracies": 1.0,
|
| 1133 |
+
"rewards/chosen": -2.7612674236297607,
|
| 1134 |
+
"rewards/margins": 17.61087989807129,
|
| 1135 |
+
"rewards/rejected": -20.372146606445312,
|
| 1136 |
+
"step": 370
|
| 1137 |
+
},
|
| 1138 |
+
{
|
| 1139 |
+
"epoch": 1.6025641025641026,
|
| 1140 |
+
"grad_norm": 0.000423431396484375,
|
| 1141 |
+
"learning_rate": 9.45177448635447e-06,
|
| 1142 |
+
"logits/chosen": -1.3942601680755615,
|
| 1143 |
+
"logits/rejected": -0.8580430746078491,
|
| 1144 |
+
"logps/chosen": -67.60791015625,
|
| 1145 |
+
"logps/rejected": -284.3832092285156,
|
| 1146 |
+
"loss": 0.0001,
|
| 1147 |
+
"rewards/accuracies": 1.0,
|
| 1148 |
+
"rewards/chosen": -2.805607795715332,
|
| 1149 |
+
"rewards/margins": 18.005619049072266,
|
| 1150 |
+
"rewards/rejected": -20.811227798461914,
|
| 1151 |
+
"step": 375
|
| 1152 |
+
},
|
| 1153 |
+
{
|
| 1154 |
+
"epoch": 1.623931623931624,
|
| 1155 |
+
"grad_norm": 0.0255126953125,
|
| 1156 |
+
"learning_rate": 9.316210990276434e-06,
|
| 1157 |
+
"logits/chosen": -1.3189040422439575,
|
| 1158 |
+
"logits/rejected": -0.8662185668945312,
|
| 1159 |
+
"logps/chosen": -67.1756362915039,
|
| 1160 |
+
"logps/rejected": -272.69500732421875,
|
| 1161 |
+
"loss": 0.0,
|
| 1162 |
+
"rewards/accuracies": 1.0,
|
| 1163 |
+
"rewards/chosen": -2.7223994731903076,
|
| 1164 |
+
"rewards/margins": 16.855377197265625,
|
| 1165 |
+
"rewards/rejected": -19.577777862548828,
|
| 1166 |
+
"step": 380
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"epoch": 1.6452991452991452,
|
| 1170 |
+
"grad_norm": 0.04638671875,
|
| 1171 |
+
"learning_rate": 9.18080542667105e-06,
|
| 1172 |
+
"logits/chosen": -1.3573819398880005,
|
| 1173 |
+
"logits/rejected": -0.8372514843940735,
|
| 1174 |
+
"logps/chosen": -68.33662414550781,
|
| 1175 |
+
"logps/rejected": -291.3061218261719,
|
| 1176 |
+
"loss": 0.0,
|
| 1177 |
+
"rewards/accuracies": 1.0,
|
| 1178 |
+
"rewards/chosen": -2.794102191925049,
|
| 1179 |
+
"rewards/margins": 18.328411102294922,
|
| 1180 |
+
"rewards/rejected": -21.122512817382812,
|
| 1181 |
+
"step": 385
|
| 1182 |
+
},
|
| 1183 |
+
{
|
| 1184 |
+
"epoch": 1.6666666666666665,
|
| 1185 |
+
"grad_norm": 0.000885009765625,
|
| 1186 |
+
"learning_rate": 9.045632889414686e-06,
|
| 1187 |
+
"logits/chosen": -1.345085859298706,
|
| 1188 |
+
"logits/rejected": -0.8467508554458618,
|
| 1189 |
+
"logps/chosen": -66.54302978515625,
|
| 1190 |
+
"logps/rejected": -283.5255126953125,
|
| 1191 |
+
"loss": 0.0001,
|
| 1192 |
+
"rewards/accuracies": 1.0,
|
| 1193 |
+
"rewards/chosen": -2.698514461517334,
|
| 1194 |
+
"rewards/margins": 17.84862518310547,
|
| 1195 |
+
"rewards/rejected": -20.54714012145996,
|
| 1196 |
+
"step": 390
|
| 1197 |
+
},
|
| 1198 |
+
{
|
| 1199 |
+
"epoch": 1.688034188034188,
|
| 1200 |
+
"grad_norm": 0.02734375,
|
| 1201 |
+
"learning_rate": 8.910768343150828e-06,
|
| 1202 |
+
"logits/chosen": -1.3259168863296509,
|
| 1203 |
+
"logits/rejected": -0.855597198009491,
|
| 1204 |
+
"logps/chosen": -69.09947204589844,
|
| 1205 |
+
"logps/rejected": -284.14801025390625,
|
| 1206 |
+
"loss": 0.0001,
|
| 1207 |
+
"rewards/accuracies": 1.0,
|
| 1208 |
+
"rewards/chosen": -3.0127596855163574,
|
| 1209 |
+
"rewards/margins": 17.691822052001953,
|
| 1210 |
+
"rewards/rejected": -20.704580307006836,
|
| 1211 |
+
"step": 395
|
| 1212 |
+
},
|
| 1213 |
+
{
|
| 1214 |
+
"epoch": 1.7094017094017095,
|
| 1215 |
+
"grad_norm": 0.0032806396484375,
|
| 1216 |
+
"learning_rate": 8.77628658171581e-06,
|
| 1217 |
+
"logits/chosen": -1.3521082401275635,
|
| 1218 |
+
"logits/rejected": -0.896456241607666,
|
| 1219 |
+
"logps/chosen": -67.49749755859375,
|
| 1220 |
+
"logps/rejected": -278.388916015625,
|
| 1221 |
+
"loss": 0.0003,
|
| 1222 |
+
"rewards/accuracies": 1.0,
|
| 1223 |
+
"rewards/chosen": -2.764543056488037,
|
| 1224 |
+
"rewards/margins": 17.227275848388672,
|
| 1225 |
+
"rewards/rejected": -19.991817474365234,
|
| 1226 |
+
"step": 400
|
| 1227 |
+
},
|
| 1228 |
+
{
|
| 1229 |
+
"epoch": 1.7307692307692308,
|
| 1230 |
+
"grad_norm": 0.0830078125,
|
| 1231 |
+
"learning_rate": 8.642262186659298e-06,
|
| 1232 |
+
"logits/chosen": -1.311095952987671,
|
| 1233 |
+
"logits/rejected": -0.8420788049697876,
|
| 1234 |
+
"logps/chosen": -68.78193664550781,
|
| 1235 |
+
"logps/rejected": -280.7834777832031,
|
| 1236 |
+
"loss": 0.0001,
|
| 1237 |
+
"rewards/accuracies": 1.0,
|
| 1238 |
+
"rewards/chosen": -2.862732172012329,
|
| 1239 |
+
"rewards/margins": 17.331756591796875,
|
| 1240 |
+
"rewards/rejected": -20.194486618041992,
|
| 1241 |
+
"step": 405
|
| 1242 |
+
},
|
| 1243 |
+
{
|
| 1244 |
+
"epoch": 1.7521367521367521,
|
| 1245 |
+
"grad_norm": 0.00640869140625,
|
| 1246 |
+
"learning_rate": 8.508769485882487e-06,
|
| 1247 |
+
"logits/chosen": -1.3232362270355225,
|
| 1248 |
+
"logits/rejected": -0.8540644645690918,
|
| 1249 |
+
"logps/chosen": -74.96504974365234,
|
| 1250 |
+
"logps/rejected": -283.9762268066406,
|
| 1251 |
+
"loss": 0.0011,
|
| 1252 |
+
"rewards/accuracies": 1.0,
|
| 1253 |
+
"rewards/chosen": -3.3426380157470703,
|
| 1254 |
+
"rewards/margins": 17.237430572509766,
|
| 1255 |
+
"rewards/rejected": -20.580068588256836,
|
| 1256 |
+
"step": 410
|
| 1257 |
+
},
|
| 1258 |
+
{
|
| 1259 |
+
"epoch": 1.7735042735042734,
|
| 1260 |
+
"grad_norm": 0.91796875,
|
| 1261 |
+
"learning_rate": 8.375882512416969e-06,
|
| 1262 |
+
"logits/chosen": -1.2893245220184326,
|
| 1263 |
+
"logits/rejected": -0.7882084846496582,
|
| 1264 |
+
"logps/chosen": -72.9918212890625,
|
| 1265 |
+
"logps/rejected": -286.07623291015625,
|
| 1266 |
+
"loss": 0.0005,
|
| 1267 |
+
"rewards/accuracies": 1.0,
|
| 1268 |
+
"rewards/chosen": -3.248950481414795,
|
| 1269 |
+
"rewards/margins": 17.547632217407227,
|
| 1270 |
+
"rewards/rejected": -20.796581268310547,
|
| 1271 |
+
"step": 415
|
| 1272 |
+
},
|
| 1273 |
+
{
|
| 1274 |
+
"epoch": 1.7948717948717947,
|
| 1275 |
+
"grad_norm": 0.019287109375,
|
| 1276 |
+
"learning_rate": 8.243674963367137e-06,
|
| 1277 |
+
"logits/chosen": -1.3166277408599854,
|
| 1278 |
+
"logits/rejected": -0.8026930093765259,
|
| 1279 |
+
"logps/chosen": -76.18501281738281,
|
| 1280 |
+
"logps/rejected": -281.90447998046875,
|
| 1281 |
+
"loss": 0.0032,
|
| 1282 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1283 |
+
"rewards/chosen": -3.600315570831299,
|
| 1284 |
+
"rewards/margins": 16.638935089111328,
|
| 1285 |
+
"rewards/rejected": -20.239248275756836,
|
| 1286 |
+
"step": 420
|
| 1287 |
+
},
|
| 1288 |
+
{
|
| 1289 |
+
"epoch": 1.8162393162393162,
|
| 1290 |
+
"grad_norm": 0.0012359619140625,
|
| 1291 |
+
"learning_rate": 8.11222015903888e-06,
|
| 1292 |
+
"logits/chosen": -1.3423035144805908,
|
| 1293 |
+
"logits/rejected": -0.8135835528373718,
|
| 1294 |
+
"logps/chosen": -78.0848159790039,
|
| 1295 |
+
"logps/rejected": -287.83135986328125,
|
| 1296 |
+
"loss": 0.0047,
|
| 1297 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1298 |
+
"rewards/chosen": -3.862813949584961,
|
| 1299 |
+
"rewards/margins": 17.16622543334961,
|
| 1300 |
+
"rewards/rejected": -21.029037475585938,
|
| 1301 |
+
"step": 425
|
| 1302 |
+
},
|
| 1303 |
+
{
|
| 1304 |
+
"epoch": 1.8376068376068377,
|
| 1305 |
+
"grad_norm": 0.037353515625,
|
| 1306 |
+
"learning_rate": 7.981591002277265e-06,
|
| 1307 |
+
"logits/chosen": -1.3140472173690796,
|
| 1308 |
+
"logits/rejected": -0.8106688261032104,
|
| 1309 |
+
"logps/chosen": -77.36860656738281,
|
| 1310 |
+
"logps/rejected": -281.99664306640625,
|
| 1311 |
+
"loss": 0.0003,
|
| 1312 |
+
"rewards/accuracies": 1.0,
|
| 1313 |
+
"rewards/chosen": -3.733330488204956,
|
| 1314 |
+
"rewards/margins": 16.830041885375977,
|
| 1315 |
+
"rewards/rejected": -20.563371658325195,
|
| 1316 |
+
"step": 430
|
| 1317 |
+
},
|
| 1318 |
+
{
|
| 1319 |
+
"epoch": 1.858974358974359,
|
| 1320 |
+
"grad_norm": 0.03515625,
|
| 1321 |
+
"learning_rate": 7.851859938035712e-06,
|
| 1322 |
+
"logits/chosen": -1.304713487625122,
|
| 1323 |
+
"logits/rejected": -0.7914744019508362,
|
| 1324 |
+
"logps/chosen": -78.41984558105469,
|
| 1325 |
+
"logps/rejected": -291.4969177246094,
|
| 1326 |
+
"loss": 0.0001,
|
| 1327 |
+
"rewards/accuracies": 1.0,
|
| 1328 |
+
"rewards/chosen": -3.7893600463867188,
|
| 1329 |
+
"rewards/margins": 17.505069732666016,
|
| 1330 |
+
"rewards/rejected": -21.294429779052734,
|
| 1331 |
+
"step": 435
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"epoch": 1.8803418803418803,
|
| 1335 |
+
"grad_norm": 0.03271484375,
|
| 1336 |
+
"learning_rate": 7.723098913199118e-06,
|
| 1337 |
+
"logits/chosen": -1.3396222591400146,
|
| 1338 |
+
"logits/rejected": -0.834884524345398,
|
| 1339 |
+
"logps/chosen": -75.61878967285156,
|
| 1340 |
+
"logps/rejected": -278.4013366699219,
|
| 1341 |
+
"loss": 0.0001,
|
| 1342 |
+
"rewards/accuracies": 1.0,
|
| 1343 |
+
"rewards/chosen": -3.660414218902588,
|
| 1344 |
+
"rewards/margins": 16.659198760986328,
|
| 1345 |
+
"rewards/rejected": -20.31961441040039,
|
| 1346 |
+
"step": 440
|
| 1347 |
+
},
|
| 1348 |
+
{
|
| 1349 |
+
"epoch": 1.9017094017094016,
|
| 1350 |
+
"grad_norm": 0.015625,
|
| 1351 |
+
"learning_rate": 7.595379336683204e-06,
|
| 1352 |
+
"logits/chosen": -1.3091070652008057,
|
| 1353 |
+
"logits/rejected": -0.7569972276687622,
|
| 1354 |
+
"logps/chosen": -72.38371276855469,
|
| 1355 |
+
"logps/rejected": -287.745849609375,
|
| 1356 |
+
"loss": 0.0,
|
| 1357 |
+
"rewards/accuracies": 1.0,
|
| 1358 |
+
"rewards/chosen": -3.400171995162964,
|
| 1359 |
+
"rewards/margins": 17.77309226989746,
|
| 1360 |
+
"rewards/rejected": -21.17326545715332,
|
| 1361 |
+
"step": 445
|
| 1362 |
+
},
|
| 1363 |
+
{
|
| 1364 |
+
"epoch": 1.9230769230769231,
|
| 1365 |
+
"grad_norm": 0.004638671875,
|
| 1366 |
+
"learning_rate": 7.468772039832218e-06,
|
| 1367 |
+
"logits/chosen": -1.2781813144683838,
|
| 1368 |
+
"logits/rejected": -0.7406023740768433,
|
| 1369 |
+
"logps/chosen": -68.62843322753906,
|
| 1370 |
+
"logps/rejected": -279.81103515625,
|
| 1371 |
+
"loss": 0.0,
|
| 1372 |
+
"rewards/accuracies": 1.0,
|
| 1373 |
+
"rewards/chosen": -3.0305323600769043,
|
| 1374 |
+
"rewards/margins": 17.346649169921875,
|
| 1375 |
+
"rewards/rejected": -20.37718391418457,
|
| 1376 |
+
"step": 450
|
| 1377 |
+
},
|
| 1378 |
+
{
|
| 1379 |
+
"epoch": 1.9444444444444444,
|
| 1380 |
+
"grad_norm": 0.0026092529296875,
|
| 1381 |
+
"learning_rate": 7.3433472371369404e-06,
|
| 1382 |
+
"logits/chosen": -1.349867820739746,
|
| 1383 |
+
"logits/rejected": -0.8363698720932007,
|
| 1384 |
+
"logps/chosen": -74.56883239746094,
|
| 1385 |
+
"logps/rejected": -285.27227783203125,
|
| 1386 |
+
"loss": 0.0,
|
| 1387 |
+
"rewards/accuracies": 1.0,
|
| 1388 |
+
"rewards/chosen": -3.506099224090576,
|
| 1389 |
+
"rewards/margins": 17.23251724243164,
|
| 1390 |
+
"rewards/rejected": -20.738616943359375,
|
| 1391 |
+
"step": 455
|
| 1392 |
+
},
|
| 1393 |
+
{
|
| 1394 |
+
"epoch": 1.965811965811966,
|
| 1395 |
+
"grad_norm": 0.00165557861328125,
|
| 1396 |
+
"learning_rate": 7.219174487294784e-06,
|
| 1397 |
+
"logits/chosen": -1.3465303182601929,
|
| 1398 |
+
"logits/rejected": -0.8472278714179993,
|
| 1399 |
+
"logps/chosen": -71.95396423339844,
|
| 1400 |
+
"logps/rejected": -288.6927795410156,
|
| 1401 |
+
"loss": 0.0001,
|
| 1402 |
+
"rewards/accuracies": 1.0,
|
| 1403 |
+
"rewards/chosen": -3.223040819168091,
|
| 1404 |
+
"rewards/margins": 17.83417510986328,
|
| 1405 |
+
"rewards/rejected": -21.05721664428711,
|
| 1406 |
+
"step": 460
|
| 1407 |
+
},
|
| 1408 |
+
{
|
| 1409 |
+
"epoch": 1.9871794871794872,
|
| 1410 |
+
"grad_norm": 0.003509521484375,
|
| 1411 |
+
"learning_rate": 7.0963226546336e-06,
|
| 1412 |
+
"logits/chosen": -1.3585379123687744,
|
| 1413 |
+
"logits/rejected": -0.8536975979804993,
|
| 1414 |
+
"logps/chosen": -72.73930358886719,
|
| 1415 |
+
"logps/rejected": -283.37579345703125,
|
| 1416 |
+
"loss": 0.0001,
|
| 1417 |
+
"rewards/accuracies": 1.0,
|
| 1418 |
+
"rewards/chosen": -3.333233594894409,
|
| 1419 |
+
"rewards/margins": 17.287670135498047,
|
| 1420 |
+
"rewards/rejected": -20.62090492248535,
|
| 1421 |
+
"step": 465
|
| 1422 |
+
},
|
| 1423 |
+
{
|
| 1424 |
+
"epoch": 1.9914529914529915,
|
| 1425 |
+
"eval_logits/chosen": -1.3740124702453613,
|
| 1426 |
+
"eval_logits/rejected": -0.8455994129180908,
|
| 1427 |
+
"eval_logps/chosen": -73.04570007324219,
|
| 1428 |
+
"eval_logps/rejected": -284.90460205078125,
|
| 1429 |
+
"eval_loss": 0.00021937819838058203,
|
| 1430 |
+
"eval_rewards/accuracies": 1.0,
|
| 1431 |
+
"eval_rewards/chosen": -3.360196590423584,
|
| 1432 |
+
"eval_rewards/margins": 17.328575134277344,
|
| 1433 |
+
"eval_rewards/rejected": -20.688772201538086,
|
| 1434 |
+
"eval_runtime": 9.5669,
|
| 1435 |
+
"eval_samples_per_second": 20.905,
|
| 1436 |
+
"eval_steps_per_second": 20.905,
|
| 1437 |
+
"step": 466
|
| 1438 |
+
},
|
| 1439 |
+
{
|
| 1440 |
+
"epoch": 2.0085470085470085,
|
| 1441 |
+
"grad_norm": 0.038818359375,
|
| 1442 |
+
"learning_rate": 6.974859870920561e-06,
|
| 1443 |
+
"logits/chosen": -1.2795295715332031,
|
| 1444 |
+
"logits/rejected": -0.8111523389816284,
|
| 1445 |
+
"logps/chosen": -75.71898651123047,
|
| 1446 |
+
"logps/rejected": -279.89599609375,
|
| 1447 |
+
"loss": 0.0,
|
| 1448 |
+
"rewards/accuracies": 1.0,
|
| 1449 |
+
"rewards/chosen": -3.6148743629455566,
|
| 1450 |
+
"rewards/margins": 16.733488082885742,
|
| 1451 |
+
"rewards/rejected": -20.348360061645508,
|
| 1452 |
+
"step": 470
|
| 1453 |
+
},
|
| 1454 |
+
{
|
| 1455 |
+
"epoch": 2.02991452991453,
|
| 1456 |
+
"grad_norm": 0.006195068359375,
|
| 1457 |
+
"learning_rate": 6.8548534975773135e-06,
|
| 1458 |
+
"logits/chosen": -1.3317922353744507,
|
| 1459 |
+
"logits/rejected": -0.8281890153884888,
|
| 1460 |
+
"logps/chosen": -75.41677856445312,
|
| 1461 |
+
"logps/rejected": -286.36761474609375,
|
| 1462 |
+
"loss": 0.0001,
|
| 1463 |
+
"rewards/accuracies": 1.0,
|
| 1464 |
+
"rewards/chosen": -3.4711709022521973,
|
| 1465 |
+
"rewards/margins": 17.307971954345703,
|
| 1466 |
+
"rewards/rejected": -20.779144287109375,
|
| 1467 |
+
"step": 475
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"epoch": 2.051282051282051,
|
| 1471 |
+
"grad_norm": 0.341796875,
|
| 1472 |
+
"learning_rate": 6.736370088322359e-06,
|
| 1473 |
+
"logits/chosen": -1.3174855709075928,
|
| 1474 |
+
"logits/rejected": -0.7978845238685608,
|
| 1475 |
+
"logps/chosen": -74.10897064208984,
|
| 1476 |
+
"logps/rejected": -283.9443664550781,
|
| 1477 |
+
"loss": 0.0004,
|
| 1478 |
+
"rewards/accuracies": 1.0,
|
| 1479 |
+
"rewards/chosen": -3.4749629497528076,
|
| 1480 |
+
"rewards/margins": 17.323604583740234,
|
| 1481 |
+
"rewards/rejected": -20.798566818237305,
|
| 1482 |
+
"step": 480
|
| 1483 |
+
},
|
| 1484 |
+
{
|
| 1485 |
+
"epoch": 2.072649572649573,
|
| 1486 |
+
"grad_norm": 0.007293701171875,
|
| 1487 |
+
"learning_rate": 6.619475352261356e-06,
|
| 1488 |
+
"logits/chosen": -1.3000952005386353,
|
| 1489 |
+
"logits/rejected": -0.8089855909347534,
|
| 1490 |
+
"logps/chosen": -78.87946319580078,
|
| 1491 |
+
"logps/rejected": -287.0957336425781,
|
| 1492 |
+
"loss": 0.0,
|
| 1493 |
+
"rewards/accuracies": 1.0,
|
| 1494 |
+
"rewards/chosen": -3.6867222785949707,
|
| 1495 |
+
"rewards/margins": 17.182767868041992,
|
| 1496 |
+
"rewards/rejected": -20.869489669799805,
|
| 1497 |
+
"step": 485
|
| 1498 |
+
},
|
| 1499 |
+
{
|
| 1500 |
+
"epoch": 2.094017094017094,
|
| 1501 |
+
"grad_norm": 0.002105712890625,
|
| 1502 |
+
"learning_rate": 6.504234117445857e-06,
|
| 1503 |
+
"logits/chosen": -1.3139859437942505,
|
| 1504 |
+
"logits/rejected": -0.8154487609863281,
|
| 1505 |
+
"logps/chosen": -74.31788635253906,
|
| 1506 |
+
"logps/rejected": -284.235107421875,
|
| 1507 |
+
"loss": 0.0,
|
| 1508 |
+
"rewards/accuracies": 1.0,
|
| 1509 |
+
"rewards/chosen": -3.5273959636688232,
|
| 1510 |
+
"rewards/margins": 17.246944427490234,
|
| 1511 |
+
"rewards/rejected": -20.774341583251953,
|
| 1512 |
+
"step": 490
|
| 1513 |
+
},
|
| 1514 |
+
{
|
| 1515 |
+
"epoch": 2.1153846153846154,
|
| 1516 |
+
"grad_norm": 0.0240478515625,
|
| 1517 |
+
"learning_rate": 6.39071029492065e-06,
|
| 1518 |
+
"logits/chosen": -1.2831732034683228,
|
| 1519 |
+
"logits/rejected": -0.7532753348350525,
|
| 1520 |
+
"logps/chosen": -73.73322296142578,
|
| 1521 |
+
"logps/rejected": -282.9715270996094,
|
| 1522 |
+
"loss": 0.0,
|
| 1523 |
+
"rewards/accuracies": 1.0,
|
| 1524 |
+
"rewards/chosen": -3.508289337158203,
|
| 1525 |
+
"rewards/margins": 17.217966079711914,
|
| 1526 |
+
"rewards/rejected": -20.726253509521484,
|
| 1527 |
+
"step": 495
|
| 1528 |
+
},
|
| 1529 |
+
{
|
| 1530 |
+
"epoch": 2.1367521367521367,
|
| 1531 |
+
"grad_norm": 0.162109375,
|
| 1532 |
+
"learning_rate": 6.2789668432796535e-06,
|
| 1533 |
+
"logits/chosen": -1.2966052293777466,
|
| 1534 |
+
"logits/rejected": -0.8182178735733032,
|
| 1535 |
+
"logps/chosen": -75.21055603027344,
|
| 1536 |
+
"logps/rejected": -284.99566650390625,
|
| 1537 |
+
"loss": 0.0001,
|
| 1538 |
+
"rewards/accuracies": 1.0,
|
| 1539 |
+
"rewards/chosen": -3.6024773120880127,
|
| 1540 |
+
"rewards/margins": 17.332172393798828,
|
| 1541 |
+
"rewards/rejected": -20.934650421142578,
|
| 1542 |
+
"step": 500
|
| 1543 |
+
},
|
| 1544 |
+
{
|
| 1545 |
+
"epoch": 2.1367521367521367,
|
| 1546 |
+
"eval_logits/chosen": -1.3704440593719482,
|
| 1547 |
+
"eval_logits/rejected": -0.8410933017730713,
|
| 1548 |
+
"eval_logps/chosen": -73.29509735107422,
|
| 1549 |
+
"eval_logps/rejected": -285.1749572753906,
|
| 1550 |
+
"eval_loss": 0.00022948597325012088,
|
| 1551 |
+
"eval_rewards/accuracies": 1.0,
|
| 1552 |
+
"eval_rewards/chosen": -3.385136365890503,
|
| 1553 |
+
"eval_rewards/margins": 17.330673217773438,
|
| 1554 |
+
"eval_rewards/rejected": -20.715810775756836,
|
| 1555 |
+
"eval_runtime": 9.5382,
|
| 1556 |
+
"eval_samples_per_second": 20.968,
|
| 1557 |
+
"eval_steps_per_second": 20.968,
|
| 1558 |
+
"step": 500
|
| 1559 |
+
},
|
| 1560 |
+
{
|
| 1561 |
+
"epoch": 2.158119658119658,
|
| 1562 |
+
"grad_norm": 0.009033203125,
|
| 1563 |
+
"learning_rate": 6.16906573375004e-06,
|
| 1564 |
+
"logits/chosen": -1.3252205848693848,
|
| 1565 |
+
"logits/rejected": -0.8390571475028992,
|
| 1566 |
+
"logps/chosen": -74.536376953125,
|
| 1567 |
+
"logps/rejected": -282.44195556640625,
|
| 1568 |
+
"loss": 0.0001,
|
| 1569 |
+
"rewards/accuracies": 1.0,
|
| 1570 |
+
"rewards/chosen": -3.4960930347442627,
|
| 1571 |
+
"rewards/margins": 17.042308807373047,
|
| 1572 |
+
"rewards/rejected": -20.538402557373047,
|
| 1573 |
+
"step": 505
|
| 1574 |
+
},
|
| 1575 |
+
{
|
| 1576 |
+
"epoch": 2.1794871794871793,
|
| 1577 |
+
"grad_norm": 0.00933837890625,
|
| 1578 |
+
"learning_rate": 6.061067915823923e-06,
|
| 1579 |
+
"logits/chosen": -1.2685729265213013,
|
| 1580 |
+
"logits/rejected": -0.7679704427719116,
|
| 1581 |
+
"logps/chosen": -72.36498260498047,
|
| 1582 |
+
"logps/rejected": -282.66351318359375,
|
| 1583 |
+
"loss": 0.0,
|
| 1584 |
+
"rewards/accuracies": 1.0,
|
| 1585 |
+
"rewards/chosen": -3.318829298019409,
|
| 1586 |
+
"rewards/margins": 17.183773040771484,
|
| 1587 |
+
"rewards/rejected": -20.50260353088379,
|
| 1588 |
+
"step": 510
|
| 1589 |
+
},
|
| 1590 |
+
{
|
| 1591 |
+
"epoch": 2.200854700854701,
|
| 1592 |
+
"grad_norm": 0.0196533203125,
|
| 1593 |
+
"learning_rate": 5.955033283456711e-06,
|
| 1594 |
+
"logits/chosen": -1.2974504232406616,
|
| 1595 |
+
"logits/rejected": -0.7774112820625305,
|
| 1596 |
+
"logps/chosen": -78.15269470214844,
|
| 1597 |
+
"logps/rejected": -293.7340393066406,
|
| 1598 |
+
"loss": 0.0001,
|
| 1599 |
+
"rewards/accuracies": 1.0,
|
| 1600 |
+
"rewards/chosen": -3.7473702430725098,
|
| 1601 |
+
"rewards/margins": 17.710281372070312,
|
| 1602 |
+
"rewards/rejected": -21.457651138305664,
|
| 1603 |
+
"step": 515
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"epoch": 2.2222222222222223,
|
| 1607 |
+
"grad_norm": 0.0023040771484375,
|
| 1608 |
+
"learning_rate": 5.8510206418507914e-06,
|
| 1609 |
+
"logits/chosen": -1.3559068441390991,
|
| 1610 |
+
"logits/rejected": -0.8591842651367188,
|
| 1611 |
+
"logps/chosen": -77.19640350341797,
|
| 1612 |
+
"logps/rejected": -300.4443359375,
|
| 1613 |
+
"loss": 0.0,
|
| 1614 |
+
"rewards/accuracies": 1.0,
|
| 1615 |
+
"rewards/chosen": -3.6723015308380127,
|
| 1616 |
+
"rewards/margins": 18.35979461669922,
|
| 1617 |
+
"rewards/rejected": -22.03209686279297,
|
| 1618 |
+
"step": 520
|
| 1619 |
+
},
|
| 1620 |
+
{
|
| 1621 |
+
"epoch": 2.2435897435897436,
|
| 1622 |
+
"grad_norm": 0.01385498046875,
|
| 1623 |
+
"learning_rate": 5.749087674843095e-06,
|
| 1624 |
+
"logits/chosen": -1.2999016046524048,
|
| 1625 |
+
"logits/rejected": -0.8241308927536011,
|
| 1626 |
+
"logps/chosen": -70.76306915283203,
|
| 1627 |
+
"logps/rejected": -283.5411376953125,
|
| 1628 |
+
"loss": 0.0,
|
| 1629 |
+
"rewards/accuracies": 1.0,
|
| 1630 |
+
"rewards/chosen": -3.232194423675537,
|
| 1631 |
+
"rewards/margins": 17.38874626159668,
|
| 1632 |
+
"rewards/rejected": -20.620941162109375,
|
| 1633 |
+
"step": 525
|
| 1634 |
+
},
|
| 1635 |
+
{
|
| 1636 |
+
"epoch": 2.264957264957265,
|
| 1637 |
+
"grad_norm": 0.0264892578125,
|
| 1638 |
+
"learning_rate": 5.649290912914482e-06,
|
| 1639 |
+
"logits/chosen": -1.3043696880340576,
|
| 1640 |
+
"logits/rejected": -0.8295344114303589,
|
| 1641 |
+
"logps/chosen": -79.15299224853516,
|
| 1642 |
+
"logps/rejected": -298.5386962890625,
|
| 1643 |
+
"loss": 0.0,
|
| 1644 |
+
"rewards/accuracies": 1.0,
|
| 1645 |
+
"rewards/chosen": -3.7782886028289795,
|
| 1646 |
+
"rewards/margins": 17.993247985839844,
|
| 1647 |
+
"rewards/rejected": -21.771535873413086,
|
| 1648 |
+
"step": 530
|
| 1649 |
+
},
|
| 1650 |
+
{
|
| 1651 |
+
"epoch": 2.286324786324786,
|
| 1652 |
+
"grad_norm": 0.04931640625,
|
| 1653 |
+
"learning_rate": 5.5516857018388144e-06,
|
| 1654 |
+
"logits/chosen": -1.355273962020874,
|
| 1655 |
+
"logits/rejected": -0.8746377229690552,
|
| 1656 |
+
"logps/chosen": -74.15048217773438,
|
| 1657 |
+
"logps/rejected": -281.2100524902344,
|
| 1658 |
+
"loss": 0.0002,
|
| 1659 |
+
"rewards/accuracies": 1.0,
|
| 1660 |
+
"rewards/chosen": -3.4310333728790283,
|
| 1661 |
+
"rewards/margins": 16.96903419494629,
|
| 1662 |
+
"rewards/rejected": -20.400066375732422,
|
| 1663 |
+
"step": 535
|
| 1664 |
+
},
|
| 1665 |
+
{
|
| 1666 |
+
"epoch": 2.3076923076923075,
|
| 1667 |
+
"grad_norm": 0.01397705078125,
|
| 1668 |
+
"learning_rate": 5.456326171989005e-06,
|
| 1669 |
+
"logits/chosen": -1.3123310804367065,
|
| 1670 |
+
"logits/rejected": -0.840388298034668,
|
| 1671 |
+
"logps/chosen": -71.68992614746094,
|
| 1672 |
+
"logps/rejected": -300.37091064453125,
|
| 1673 |
+
"loss": 0.0,
|
| 1674 |
+
"rewards/accuracies": 1.0,
|
| 1675 |
+
"rewards/chosen": -3.202878475189209,
|
| 1676 |
+
"rewards/margins": 18.644420623779297,
|
| 1677 |
+
"rewards/rejected": -21.84729766845703,
|
| 1678 |
+
"step": 540
|
| 1679 |
+
},
|
| 1680 |
+
{
|
| 1681 |
+
"epoch": 2.3290598290598292,
|
| 1682 |
+
"grad_norm": 0.0027618408203125,
|
| 1683 |
+
"learning_rate": 5.363265208317156e-06,
|
| 1684 |
+
"logits/chosen": -1.2788275480270386,
|
| 1685 |
+
"logits/rejected": -0.8199743032455444,
|
| 1686 |
+
"logps/chosen": -73.48957824707031,
|
| 1687 |
+
"logps/rejected": -281.1923828125,
|
| 1688 |
+
"loss": 0.0,
|
| 1689 |
+
"rewards/accuracies": 1.0,
|
| 1690 |
+
"rewards/chosen": -3.415186643600464,
|
| 1691 |
+
"rewards/margins": 17.106266021728516,
|
| 1692 |
+
"rewards/rejected": -20.52145004272461,
|
| 1693 |
+
"step": 545
|
| 1694 |
+
},
|
| 1695 |
+
{
|
| 1696 |
+
"epoch": 2.3504273504273505,
|
| 1697 |
+
"grad_norm": 0.01055908203125,
|
| 1698 |
+
"learning_rate": 5.272554421025347e-06,
|
| 1699 |
+
"logits/chosen": -1.3188756704330444,
|
| 1700 |
+
"logits/rejected": -0.8151782751083374,
|
| 1701 |
+
"logps/chosen": -74.64764404296875,
|
| 1702 |
+
"logps/rejected": -291.92108154296875,
|
| 1703 |
+
"loss": 0.0,
|
| 1704 |
+
"rewards/accuracies": 1.0,
|
| 1705 |
+
"rewards/chosen": -3.5020947456359863,
|
| 1706 |
+
"rewards/margins": 17.861783981323242,
|
| 1707 |
+
"rewards/rejected": -21.36387825012207,
|
| 1708 |
+
"step": 550
|
| 1709 |
+
},
|
| 1710 |
+
{
|
| 1711 |
+
"epoch": 2.371794871794872,
|
| 1712 |
+
"grad_norm": 0.05078125,
|
| 1713 |
+
"learning_rate": 5.184244116943411e-06,
|
| 1714 |
+
"logits/chosen": -1.3126680850982666,
|
| 1715 |
+
"logits/rejected": -0.8074380159378052,
|
| 1716 |
+
"logps/chosen": -73.89201354980469,
|
| 1717 |
+
"logps/rejected": -285.258056640625,
|
| 1718 |
+
"loss": 0.003,
|
| 1719 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1720 |
+
"rewards/chosen": -3.3727481365203857,
|
| 1721 |
+
"rewards/margins": 17.37813377380371,
|
| 1722 |
+
"rewards/rejected": -20.75088119506836,
|
| 1723 |
+
"step": 555
|
| 1724 |
+
},
|
| 1725 |
+
{
|
| 1726 |
+
"epoch": 2.393162393162393,
|
| 1727 |
+
"grad_norm": 0.00604248046875,
|
| 1728 |
+
"learning_rate": 5.098383271629512e-06,
|
| 1729 |
+
"logits/chosen": -1.3314543962478638,
|
| 1730 |
+
"logits/rejected": -0.8163145184516907,
|
| 1731 |
+
"logps/chosen": -73.50102233886719,
|
| 1732 |
+
"logps/rejected": -279.52532958984375,
|
| 1733 |
+
"loss": 0.0001,
|
| 1734 |
+
"rewards/accuracies": 1.0,
|
| 1735 |
+
"rewards/chosen": -3.414262294769287,
|
| 1736 |
+
"rewards/margins": 16.943660736083984,
|
| 1737 |
+
"rewards/rejected": -20.357921600341797,
|
| 1738 |
+
"step": 560
|
| 1739 |
+
},
|
| 1740 |
+
{
|
| 1741 |
+
"epoch": 2.4145299145299144,
|
| 1742 |
+
"grad_norm": 0.00311279296875,
|
| 1743 |
+
"learning_rate": 5.015019502209056e-06,
|
| 1744 |
+
"logits/chosen": -1.3196806907653809,
|
| 1745 |
+
"logits/rejected": -0.8105131387710571,
|
| 1746 |
+
"logps/chosen": -72.18685913085938,
|
| 1747 |
+
"logps/rejected": -275.95123291015625,
|
| 1748 |
+
"loss": 0.0001,
|
| 1749 |
+
"rewards/accuracies": 1.0,
|
| 1750 |
+
"rewards/chosen": -3.361246109008789,
|
| 1751 |
+
"rewards/margins": 16.643529891967773,
|
| 1752 |
+
"rewards/rejected": -20.004776000976562,
|
| 1753 |
+
"step": 565
|
| 1754 |
+
},
|
| 1755 |
+
{
|
| 1756 |
+
"epoch": 2.435897435897436,
|
| 1757 |
+
"grad_norm": 0.000823974609375,
|
| 1758 |
+
"learning_rate": 4.934199040966955e-06,
|
| 1759 |
+
"logits/chosen": -1.3401740789413452,
|
| 1760 |
+
"logits/rejected": -0.8449984788894653,
|
| 1761 |
+
"logps/chosen": -73.99894714355469,
|
| 1762 |
+
"logps/rejected": -279.31915283203125,
|
| 1763 |
+
"loss": 0.0001,
|
| 1764 |
+
"rewards/accuracies": 1.0,
|
| 1765 |
+
"rewards/chosen": -3.482757568359375,
|
| 1766 |
+
"rewards/margins": 16.8636531829834,
|
| 1767 |
+
"rewards/rejected": -20.346412658691406,
|
| 1768 |
+
"step": 570
|
| 1769 |
+
},
|
| 1770 |
+
{
|
| 1771 |
+
"epoch": 2.4572649572649574,
|
| 1772 |
+
"grad_norm": 0.03271484375,
|
| 1773 |
+
"learning_rate": 4.855966709707881e-06,
|
| 1774 |
+
"logits/chosen": -1.308977484703064,
|
| 1775 |
+
"logits/rejected": -0.8370776176452637,
|
| 1776 |
+
"logps/chosen": -77.53469848632812,
|
| 1777 |
+
"logps/rejected": -285.58367919921875,
|
| 1778 |
+
"loss": 0.0029,
|
| 1779 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1780 |
+
"rewards/chosen": -3.7107481956481934,
|
| 1781 |
+
"rewards/margins": 17.121841430664062,
|
| 1782 |
+
"rewards/rejected": -20.832592010498047,
|
| 1783 |
+
"step": 575
|
| 1784 |
+
},
|
| 1785 |
+
{
|
| 1786 |
+
"epoch": 2.4786324786324787,
|
| 1787 |
+
"grad_norm": 0.0299072265625,
|
| 1788 |
+
"learning_rate": 4.780365894898799e-06,
|
| 1789 |
+
"logits/chosen": -1.3271667957305908,
|
| 1790 |
+
"logits/rejected": -0.8259018063545227,
|
| 1791 |
+
"logps/chosen": -74.68269348144531,
|
| 1792 |
+
"logps/rejected": -287.12078857421875,
|
| 1793 |
+
"loss": 0.0,
|
| 1794 |
+
"rewards/accuracies": 1.0,
|
| 1795 |
+
"rewards/chosen": -3.4662697315216064,
|
| 1796 |
+
"rewards/margins": 17.42196273803711,
|
| 1797 |
+
"rewards/rejected": -20.888233184814453,
|
| 1798 |
+
"step": 580
|
| 1799 |
+
},
|
| 1800 |
+
{
|
| 1801 |
+
"epoch": 2.5,
|
| 1802 |
+
"grad_norm": 0.00921630859375,
|
| 1803 |
+
"learning_rate": 4.7074385236074684e-06,
|
| 1804 |
+
"logits/chosen": -1.3541457653045654,
|
| 1805 |
+
"logits/rejected": -0.8319869041442871,
|
| 1806 |
+
"logps/chosen": -78.34286499023438,
|
| 1807 |
+
"logps/rejected": -294.329833984375,
|
| 1808 |
+
"loss": 0.0,
|
| 1809 |
+
"rewards/accuracies": 1.0,
|
| 1810 |
+
"rewards/chosen": -3.6850643157958984,
|
| 1811 |
+
"rewards/margins": 17.795787811279297,
|
| 1812 |
+
"rewards/rejected": -21.480854034423828,
|
| 1813 |
+
"step": 585
|
| 1814 |
+
},
|
| 1815 |
+
{
|
| 1816 |
+
"epoch": 2.5213675213675213,
|
| 1817 |
+
"grad_norm": 0.052001953125,
|
| 1818 |
+
"learning_rate": 4.63722504025034e-06,
|
| 1819 |
+
"logits/chosen": -1.3320066928863525,
|
| 1820 |
+
"logits/rejected": -0.8415569067001343,
|
| 1821 |
+
"logps/chosen": -71.98558044433594,
|
| 1822 |
+
"logps/rejected": -284.99346923828125,
|
| 1823 |
+
"loss": 0.0011,
|
| 1824 |
+
"rewards/accuracies": 1.0,
|
| 1825 |
+
"rewards/chosen": -3.249178647994995,
|
| 1826 |
+
"rewards/margins": 17.486406326293945,
|
| 1827 |
+
"rewards/rejected": -20.735586166381836,
|
| 1828 |
+
"step": 590
|
| 1829 |
+
},
|
| 1830 |
+
{
|
| 1831 |
+
"epoch": 2.5427350427350426,
|
| 1832 |
+
"grad_norm": 0.01025390625,
|
| 1833 |
+
"learning_rate": 4.569764384162676e-06,
|
| 1834 |
+
"logits/chosen": -1.3463108539581299,
|
| 1835 |
+
"logits/rejected": -0.8353781700134277,
|
| 1836 |
+
"logps/chosen": -68.05410766601562,
|
| 1837 |
+
"logps/rejected": -284.91461181640625,
|
| 1838 |
+
"loss": 0.0001,
|
| 1839 |
+
"rewards/accuracies": 1.0,
|
| 1840 |
+
"rewards/chosen": -2.9681944847106934,
|
| 1841 |
+
"rewards/margins": 17.779176712036133,
|
| 1842 |
+
"rewards/rejected": -20.74736976623535,
|
| 1843 |
+
"step": 595
|
| 1844 |
+
},
|
| 1845 |
+
{
|
| 1846 |
+
"epoch": 2.564102564102564,
|
| 1847 |
+
"grad_norm": 0.01422119140625,
|
| 1848 |
+
"learning_rate": 4.50509396800341e-06,
|
| 1849 |
+
"logits/chosen": -1.2894313335418701,
|
| 1850 |
+
"logits/rejected": -0.7797183990478516,
|
| 1851 |
+
"logps/chosen": -72.90419006347656,
|
| 1852 |
+
"logps/rejected": -285.7866516113281,
|
| 1853 |
+
"loss": 0.0,
|
| 1854 |
+
"rewards/accuracies": 1.0,
|
| 1855 |
+
"rewards/chosen": -3.3224644660949707,
|
| 1856 |
+
"rewards/margins": 17.589801788330078,
|
| 1857 |
+
"rewards/rejected": -20.91226577758789,
|
| 1858 |
+
"step": 600
|
| 1859 |
+
},
|
| 1860 |
+
{
|
| 1861 |
+
"epoch": 2.5854700854700856,
|
| 1862 |
+
"grad_norm": 0.0247802734375,
|
| 1863 |
+
"learning_rate": 4.443249657006627e-06,
|
| 1864 |
+
"logits/chosen": -1.2982523441314697,
|
| 1865 |
+
"logits/rejected": -0.7844825983047485,
|
| 1866 |
+
"logps/chosen": -70.4168472290039,
|
| 1867 |
+
"logps/rejected": -290.3528747558594,
|
| 1868 |
+
"loss": 0.0029,
|
| 1869 |
+
"rewards/accuracies": 0.9958333969116211,
|
| 1870 |
+
"rewards/chosen": -3.1497464179992676,
|
| 1871 |
+
"rewards/margins": 18.14370346069336,
|
| 1872 |
+
"rewards/rejected": -21.293447494506836,
|
| 1873 |
+
"step": 605
|
| 1874 |
+
},
|
| 1875 |
+
{
|
| 1876 |
+
"epoch": 2.606837606837607,
|
| 1877 |
+
"grad_norm": 0.0164794921875,
|
| 1878 |
+
"learning_rate": 4.384265749091266e-06,
|
| 1879 |
+
"logits/chosen": -1.2762781381607056,
|
| 1880 |
+
"logits/rejected": -0.7862453460693359,
|
| 1881 |
+
"logps/chosen": -78.71661376953125,
|
| 1882 |
+
"logps/rejected": -288.5256042480469,
|
| 1883 |
+
"loss": 0.0001,
|
| 1884 |
+
"rewards/accuracies": 1.0,
|
| 1885 |
+
"rewards/chosen": -3.7849228382110596,
|
| 1886 |
+
"rewards/margins": 17.296167373657227,
|
| 1887 |
+
"rewards/rejected": -21.081090927124023,
|
| 1888 |
+
"step": 610
|
| 1889 |
+
},
|
| 1890 |
+
{
|
| 1891 |
+
"epoch": 2.628205128205128,
|
| 1892 |
+
"grad_norm": 0.00482177734375,
|
| 1893 |
+
"learning_rate": 4.328174955840002e-06,
|
| 1894 |
+
"logits/chosen": -1.2989494800567627,
|
| 1895 |
+
"logits/rejected": -0.791740357875824,
|
| 1896 |
+
"logps/chosen": -68.19273376464844,
|
| 1897 |
+
"logps/rejected": -283.67010498046875,
|
| 1898 |
+
"loss": 0.0,
|
| 1899 |
+
"rewards/accuracies": 1.0,
|
| 1900 |
+
"rewards/chosen": -3.0757412910461426,
|
| 1901 |
+
"rewards/margins": 17.688796997070312,
|
| 1902 |
+
"rewards/rejected": -20.764535903930664,
|
| 1903 |
+
"step": 615
|
| 1904 |
+
},
|
| 1905 |
+
{
|
| 1906 |
+
"epoch": 2.6495726495726495,
|
| 1907 |
+
"grad_norm": 0.0026092529296875,
|
| 1908 |
+
"learning_rate": 4.275008384357902e-06,
|
| 1909 |
+
"logits/chosen": -1.3389320373535156,
|
| 1910 |
+
"logits/rejected": -0.8387205004692078,
|
| 1911 |
+
"logps/chosen": -72.53665924072266,
|
| 1912 |
+
"logps/rejected": -283.5371398925781,
|
| 1913 |
+
"loss": 0.0,
|
| 1914 |
+
"rewards/accuracies": 1.0,
|
| 1915 |
+
"rewards/chosen": -3.2912750244140625,
|
| 1916 |
+
"rewards/margins": 17.413835525512695,
|
| 1917 |
+
"rewards/rejected": -20.705108642578125,
|
| 1918 |
+
"step": 620
|
| 1919 |
+
},
|
| 1920 |
+
{
|
| 1921 |
+
"epoch": 2.6709401709401708,
|
| 1922 |
+
"grad_norm": 0.0028533935546875,
|
| 1923 |
+
"learning_rate": 4.224795520020898e-06,
|
| 1924 |
+
"logits/chosen": -1.2840917110443115,
|
| 1925 |
+
"logits/rejected": -0.7634187936782837,
|
| 1926 |
+
"logps/chosen": -75.63284301757812,
|
| 1927 |
+
"logps/rejected": -282.2013244628906,
|
| 1928 |
+
"loss": 0.0001,
|
| 1929 |
+
"rewards/accuracies": 1.0,
|
| 1930 |
+
"rewards/chosen": -3.4395980834960938,
|
| 1931 |
+
"rewards/margins": 16.99862289428711,
|
| 1932 |
+
"rewards/rejected": -20.438220977783203,
|
| 1933 |
+
"step": 625
|
| 1934 |
+
},
|
| 1935 |
+
{
|
| 1936 |
+
"epoch": 2.6923076923076925,
|
| 1937 |
+
"grad_norm": 0.01446533203125,
|
| 1938 |
+
"learning_rate": 4.177564210123634e-06,
|
| 1939 |
+
"logits/chosen": -1.32615327835083,
|
| 1940 |
+
"logits/rejected": -0.8317953944206238,
|
| 1941 |
+
"logps/chosen": -72.77333068847656,
|
| 1942 |
+
"logps/rejected": -291.9281921386719,
|
| 1943 |
+
"loss": 0.0,
|
| 1944 |
+
"rewards/accuracies": 1.0,
|
| 1945 |
+
"rewards/chosen": -3.310988664627075,
|
| 1946 |
+
"rewards/margins": 17.982282638549805,
|
| 1947 |
+
"rewards/rejected": -21.293270111083984,
|
| 1948 |
+
"step": 630
|
| 1949 |
+
},
|
| 1950 |
+
{
|
| 1951 |
+
"epoch": 2.713675213675214,
|
| 1952 |
+
"grad_norm": 0.00921630859375,
|
| 1953 |
+
"learning_rate": 4.133340648435789e-06,
|
| 1954 |
+
"logits/chosen": -1.3119795322418213,
|
| 1955 |
+
"logits/rejected": -0.7846705913543701,
|
| 1956 |
+
"logps/chosen": -74.24435424804688,
|
| 1957 |
+
"logps/rejected": -289.03887939453125,
|
| 1958 |
+
"loss": 0.0,
|
| 1959 |
+
"rewards/accuracies": 1.0,
|
| 1960 |
+
"rewards/chosen": -3.4849464893341064,
|
| 1961 |
+
"rewards/margins": 17.662567138671875,
|
| 1962 |
+
"rewards/rejected": -21.147512435913086,
|
| 1963 |
+
"step": 635
|
| 1964 |
+
},
|
| 1965 |
+
{
|
| 1966 |
+
"epoch": 2.735042735042735,
|
| 1967 |
+
"grad_norm": 0.005828857421875,
|
| 1968 |
+
"learning_rate": 4.092149360675402e-06,
|
| 1969 |
+
"logits/chosen": -1.2881155014038086,
|
| 1970 |
+
"logits/rejected": -0.8033782243728638,
|
| 1971 |
+
"logps/chosen": -79.22930908203125,
|
| 1972 |
+
"logps/rejected": -294.1476745605469,
|
| 1973 |
+
"loss": 0.0,
|
| 1974 |
+
"rewards/accuracies": 1.0,
|
| 1975 |
+
"rewards/chosen": -3.7118358612060547,
|
| 1976 |
+
"rewards/margins": 17.629985809326172,
|
| 1977 |
+
"rewards/rejected": -21.341821670532227,
|
| 1978 |
+
"step": 640
|
| 1979 |
+
},
|
| 1980 |
+
{
|
| 1981 |
+
"epoch": 2.7564102564102564,
|
| 1982 |
+
"grad_norm": 0.019775390625,
|
| 1983 |
+
"learning_rate": 4.054013190907282e-06,
|
| 1984 |
+
"logits/chosen": -1.2686903476715088,
|
| 1985 |
+
"logits/rejected": -0.7805891633033752,
|
| 1986 |
+
"logps/chosen": -70.22049713134766,
|
| 1987 |
+
"logps/rejected": -284.9203796386719,
|
| 1988 |
+
"loss": 0.0,
|
| 1989 |
+
"rewards/accuracies": 1.0,
|
| 1990 |
+
"rewards/chosen": -3.1786301136016846,
|
| 1991 |
+
"rewards/margins": 17.73773956298828,
|
| 1992 |
+
"rewards/rejected": -20.916370391845703,
|
| 1993 |
+
"step": 645
|
| 1994 |
+
},
|
| 1995 |
+
{
|
| 1996 |
+
"epoch": 2.7777777777777777,
|
| 1997 |
+
"grad_norm": 0.004241943359375,
|
| 1998 |
+
"learning_rate": 4.018953288874035e-06,
|
| 1999 |
+
"logits/chosen": -1.3032779693603516,
|
| 2000 |
+
"logits/rejected": -0.8233755230903625,
|
| 2001 |
+
"logps/chosen": -74.59503936767578,
|
| 2002 |
+
"logps/rejected": -288.961669921875,
|
| 2003 |
+
"loss": 0.0,
|
| 2004 |
+
"rewards/accuracies": 1.0,
|
| 2005 |
+
"rewards/chosen": -3.4693355560302734,
|
| 2006 |
+
"rewards/margins": 17.675161361694336,
|
| 2007 |
+
"rewards/rejected": -21.14449691772461,
|
| 2008 |
+
"step": 650
|
| 2009 |
+
},
|
| 2010 |
+
{
|
| 2011 |
+
"epoch": 2.799145299145299,
|
| 2012 |
+
"grad_norm": 0.0040283203125,
|
| 2013 |
+
"learning_rate": 3.9869890982667385e-06,
|
| 2014 |
+
"logits/chosen": -1.310773491859436,
|
| 2015 |
+
"logits/rejected": -0.7524069547653198,
|
| 2016 |
+
"logps/chosen": -72.80381774902344,
|
| 2017 |
+
"logps/rejected": -290.0320739746094,
|
| 2018 |
+
"loss": 0.0,
|
| 2019 |
+
"rewards/accuracies": 1.0,
|
| 2020 |
+
"rewards/chosen": -3.345860004425049,
|
| 2021 |
+
"rewards/margins": 17.688016891479492,
|
| 2022 |
+
"rewards/rejected": -21.033876419067383,
|
| 2023 |
+
"step": 655
|
| 2024 |
+
},
|
| 2025 |
+
{
|
| 2026 |
+
"epoch": 2.8205128205128203,
|
| 2027 |
+
"grad_norm": 0.0189208984375,
|
| 2028 |
+
"learning_rate": 3.9581383459417625e-06,
|
| 2029 |
+
"logits/chosen": -1.291512370109558,
|
| 2030 |
+
"logits/rejected": -0.790591299533844,
|
| 2031 |
+
"logps/chosen": -78.11724853515625,
|
| 2032 |
+
"logps/rejected": -297.84283447265625,
|
| 2033 |
+
"loss": 0.0001,
|
| 2034 |
+
"rewards/accuracies": 1.0,
|
| 2035 |
+
"rewards/chosen": -3.7806007862091064,
|
| 2036 |
+
"rewards/margins": 18.05144500732422,
|
| 2037 |
+
"rewards/rejected": -21.832046508789062,
|
| 2038 |
+
"step": 660
|
| 2039 |
+
},
|
| 2040 |
+
{
|
| 2041 |
+
"epoch": 2.841880341880342,
|
| 2042 |
+
"grad_norm": 0.00494384765625,
|
| 2043 |
+
"learning_rate": 3.932417032089722e-06,
|
| 2044 |
+
"logits/chosen": -1.3292133808135986,
|
| 2045 |
+
"logits/rejected": -0.8189595937728882,
|
| 2046 |
+
"logps/chosen": -76.23522186279297,
|
| 2047 |
+
"logps/rejected": -293.2325744628906,
|
| 2048 |
+
"loss": 0.0001,
|
| 2049 |
+
"rewards/accuracies": 1.0,
|
| 2050 |
+
"rewards/chosen": -3.59516978263855,
|
| 2051 |
+
"rewards/margins": 17.823274612426758,
|
| 2052 |
+
"rewards/rejected": -21.418445587158203,
|
| 2053 |
+
"step": 665
|
| 2054 |
+
},
|
| 2055 |
+
{
|
| 2056 |
+
"epoch": 2.8632478632478633,
|
| 2057 |
+
"grad_norm": 0.0029754638671875,
|
| 2058 |
+
"learning_rate": 3.909839421362017e-06,
|
| 2059 |
+
"logits/chosen": -1.2779964208602905,
|
| 2060 |
+
"logits/rejected": -0.7794166803359985,
|
| 2061 |
+
"logps/chosen": -74.2072982788086,
|
| 2062 |
+
"logps/rejected": -289.36956787109375,
|
| 2063 |
+
"loss": 0.0,
|
| 2064 |
+
"rewards/accuracies": 1.0,
|
| 2065 |
+
"rewards/chosen": -3.46783185005188,
|
| 2066 |
+
"rewards/margins": 17.554330825805664,
|
| 2067 |
+
"rewards/rejected": -21.02216148376465,
|
| 2068 |
+
"step": 670
|
| 2069 |
+
},
|
| 2070 |
+
{
|
| 2071 |
+
"epoch": 2.8846153846153846,
|
| 2072 |
+
"grad_norm": 0.038818359375,
|
| 2073 |
+
"learning_rate": 3.890418034959871e-06,
|
| 2074 |
+
"logits/chosen": -1.2737759351730347,
|
| 2075 |
+
"logits/rejected": -0.7384223937988281,
|
| 2076 |
+
"logps/chosen": -73.76658630371094,
|
| 2077 |
+
"logps/rejected": -286.06195068359375,
|
| 2078 |
+
"loss": 0.0,
|
| 2079 |
+
"rewards/accuracies": 1.0,
|
| 2080 |
+
"rewards/chosen": -3.443812847137451,
|
| 2081 |
+
"rewards/margins": 17.49923324584961,
|
| 2082 |
+
"rewards/rejected": -20.943044662475586,
|
| 2083 |
+
"step": 675
|
| 2084 |
+
},
|
| 2085 |
+
{
|
| 2086 |
+
"epoch": 2.905982905982906,
|
| 2087 |
+
"grad_norm": 0.06884765625,
|
| 2088 |
+
"learning_rate": 3.874163643690263e-06,
|
| 2089 |
+
"logits/chosen": -1.255707025527954,
|
| 2090 |
+
"logits/rejected": -0.7339369654655457,
|
| 2091 |
+
"logps/chosen": -80.28028869628906,
|
| 2092 |
+
"logps/rejected": -291.90179443359375,
|
| 2093 |
+
"loss": 0.0,
|
| 2094 |
+
"rewards/accuracies": 1.0,
|
| 2095 |
+
"rewards/chosen": -3.9126086235046387,
|
| 2096 |
+
"rewards/margins": 17.456329345703125,
|
| 2097 |
+
"rewards/rejected": -21.368938446044922,
|
| 2098 |
+
"step": 680
|
| 2099 |
+
},
|
| 2100 |
+
{
|
| 2101 |
+
"epoch": 2.9273504273504276,
|
| 2102 |
+
"grad_norm": 0.04296875,
|
| 2103 |
+
"learning_rate": 3.861085261992599e-06,
|
| 2104 |
+
"logits/chosen": -1.306028127670288,
|
| 2105 |
+
"logits/rejected": -0.8490394353866577,
|
| 2106 |
+
"logps/chosen": -78.53582763671875,
|
| 2107 |
+
"logps/rejected": -289.38330078125,
|
| 2108 |
+
"loss": 0.0001,
|
| 2109 |
+
"rewards/accuracies": 1.0,
|
| 2110 |
+
"rewards/chosen": -3.830167770385742,
|
| 2111 |
+
"rewards/margins": 17.217695236206055,
|
| 2112 |
+
"rewards/rejected": -21.047863006591797,
|
| 2113 |
+
"step": 685
|
| 2114 |
+
},
|
| 2115 |
+
{
|
| 2116 |
+
"epoch": 2.948717948717949,
|
| 2117 |
+
"grad_norm": 0.12255859375,
|
| 2118 |
+
"learning_rate": 3.851190142939442e-06,
|
| 2119 |
+
"logits/chosen": -1.2999136447906494,
|
| 2120 |
+
"logits/rejected": -0.8062965273857117,
|
| 2121 |
+
"logps/chosen": -72.80134582519531,
|
| 2122 |
+
"logps/rejected": -288.02532958984375,
|
| 2123 |
+
"loss": 0.0001,
|
| 2124 |
+
"rewards/accuracies": 1.0,
|
| 2125 |
+
"rewards/chosen": -3.2750275135040283,
|
| 2126 |
+
"rewards/margins": 17.672157287597656,
|
| 2127 |
+
"rewards/rejected": -20.94718360900879,
|
| 2128 |
+
"step": 690
|
| 2129 |
+
},
|
| 2130 |
+
{
|
| 2131 |
+
"epoch": 2.97008547008547,
|
| 2132 |
+
"grad_norm": 0.001434326171875,
|
| 2133 |
+
"learning_rate": 3.844483774214069e-06,
|
| 2134 |
+
"logits/chosen": -1.2856634855270386,
|
| 2135 |
+
"logits/rejected": -0.7375695705413818,
|
| 2136 |
+
"logps/chosen": -72.40345764160156,
|
| 2137 |
+
"logps/rejected": -288.3551940917969,
|
| 2138 |
+
"loss": 0.0003,
|
| 2139 |
+
"rewards/accuracies": 1.0,
|
| 2140 |
+
"rewards/chosen": -3.333986759185791,
|
| 2141 |
+
"rewards/margins": 17.785888671875,
|
| 2142 |
+
"rewards/rejected": -21.119874954223633,
|
| 2143 |
+
"step": 695
|
| 2144 |
+
},
|
| 2145 |
+
{
|
| 2146 |
+
"epoch": 2.9871794871794872,
|
| 2147 |
+
"eval_logits/chosen": -1.3524901866912842,
|
| 2148 |
+
"eval_logits/rejected": -0.8189607262611389,
|
| 2149 |
+
"eval_logps/chosen": -73.21478271484375,
|
| 2150 |
+
"eval_logps/rejected": -285.69122314453125,
|
| 2151 |
+
"eval_loss": 0.00020370040147099644,
|
| 2152 |
+
"eval_rewards/accuracies": 1.0,
|
| 2153 |
+
"eval_rewards/chosen": -3.377105474472046,
|
| 2154 |
+
"eval_rewards/margins": 17.390329360961914,
|
| 2155 |
+
"eval_rewards/rejected": -20.767436981201172,
|
| 2156 |
+
"eval_runtime": 9.5425,
|
| 2157 |
+
"eval_samples_per_second": 20.959,
|
| 2158 |
+
"eval_steps_per_second": 20.959,
|
| 2159 |
+
"step": 699
|
| 2160 |
+
}
|
| 2161 |
+
],
|
| 2162 |
+
"logging_steps": 5,
|
| 2163 |
+
"max_steps": 702,
|
| 2164 |
+
"num_input_tokens_seen": 0,
|
| 2165 |
+
"num_train_epochs": 3,
|
| 2166 |
+
"save_steps": 500,
|
| 2167 |
+
"stateful_callbacks": {
|
| 2168 |
+
"TrainerControl": {
|
| 2169 |
+
"args": {
|
| 2170 |
+
"should_epoch_stop": false,
|
| 2171 |
+
"should_evaluate": false,
|
| 2172 |
+
"should_log": false,
|
| 2173 |
+
"should_save": true,
|
| 2174 |
+
"should_training_stop": false
|
| 2175 |
+
},
|
| 2176 |
+
"attributes": {}
|
| 2177 |
+
}
|
| 2178 |
+
},
|
| 2179 |
+
"total_flos": 0.0,
|
| 2180 |
+
"train_batch_size": 12,
|
| 2181 |
+
"trial_name": null,
|
| 2182 |
+
"trial_params": null
|
| 2183 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8ecb6305a7f4e8192099c5fceaacdcaa3bcd58361c2ddb55ea68c16d4d768637
|
| 3 |
+
size 6520
|