Upload folder using huggingface_hub
Browse files- .gitattributes +1 -0
- added_tokens.json +28 -0
- chat_template.jinja +89 -0
- config.json +30 -0
- generation_config.json +13 -0
- merges.txt +0 -0
- model.safetensors +3 -0
- special_tokens_map.json +31 -0
- tokenizer.json +3 -0
- tokenizer_config.json +239 -0
- trainer_state.json +1734 -0
- training_args.bin +3 -0
- vocab.json +0 -0
.gitattributes
CHANGED
|
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
|
|
|
|
|
| 33 |
*.zip filter=lfs diff=lfs merge=lfs -text
|
| 34 |
*.zst filter=lfs diff=lfs merge=lfs -text
|
| 35 |
*tfevents* filter=lfs diff=lfs merge=lfs -text
|
| 36 |
+
tokenizer.json filter=lfs diff=lfs merge=lfs -text
|
added_tokens.json
ADDED
|
@@ -0,0 +1,28 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"</think>": 151668,
|
| 3 |
+
"</tool_call>": 151658,
|
| 4 |
+
"</tool_response>": 151666,
|
| 5 |
+
"<think>": 151667,
|
| 6 |
+
"<tool_call>": 151657,
|
| 7 |
+
"<tool_response>": 151665,
|
| 8 |
+
"<|box_end|>": 151649,
|
| 9 |
+
"<|box_start|>": 151648,
|
| 10 |
+
"<|endoftext|>": 151643,
|
| 11 |
+
"<|file_sep|>": 151664,
|
| 12 |
+
"<|fim_middle|>": 151660,
|
| 13 |
+
"<|fim_pad|>": 151662,
|
| 14 |
+
"<|fim_prefix|>": 151659,
|
| 15 |
+
"<|fim_suffix|>": 151661,
|
| 16 |
+
"<|im_end|>": 151645,
|
| 17 |
+
"<|im_start|>": 151644,
|
| 18 |
+
"<|image_pad|>": 151655,
|
| 19 |
+
"<|object_ref_end|>": 151647,
|
| 20 |
+
"<|object_ref_start|>": 151646,
|
| 21 |
+
"<|quad_end|>": 151651,
|
| 22 |
+
"<|quad_start|>": 151650,
|
| 23 |
+
"<|repo_name|>": 151663,
|
| 24 |
+
"<|video_pad|>": 151656,
|
| 25 |
+
"<|vision_end|>": 151653,
|
| 26 |
+
"<|vision_pad|>": 151654,
|
| 27 |
+
"<|vision_start|>": 151652
|
| 28 |
+
}
|
chat_template.jinja
ADDED
|
@@ -0,0 +1,89 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{%- if tools %}
|
| 2 |
+
{{- '<|im_start|>system\n' }}
|
| 3 |
+
{%- if messages[0].role == 'system' %}
|
| 4 |
+
{{- messages[0].content + '\n\n' }}
|
| 5 |
+
{%- endif %}
|
| 6 |
+
{{- "# Tools\n\nYou may call one or more functions to assist with the user query.\n\nYou are provided with function signatures within <tools></tools> XML tags:\n<tools>" }}
|
| 7 |
+
{%- for tool in tools %}
|
| 8 |
+
{{- "\n" }}
|
| 9 |
+
{{- tool | tojson }}
|
| 10 |
+
{%- endfor %}
|
| 11 |
+
{{- "\n</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call><|im_end|>\n" }}
|
| 12 |
+
{%- else %}
|
| 13 |
+
{%- if messages[0].role == 'system' %}
|
| 14 |
+
{{- '<|im_start|>system\n' + messages[0].content + '<|im_end|>\n' }}
|
| 15 |
+
{%- endif %}
|
| 16 |
+
{%- endif %}
|
| 17 |
+
{%- set ns = namespace(multi_step_tool=true, last_query_index=messages|length - 1) %}
|
| 18 |
+
{%- for message in messages[::-1] %}
|
| 19 |
+
{%- set index = (messages|length - 1) - loop.index0 %}
|
| 20 |
+
{%- if ns.multi_step_tool and message.role == "user" and message.content is string and not(message.content.startswith('<tool_response>') and message.content.endswith('</tool_response>')) %}
|
| 21 |
+
{%- set ns.multi_step_tool = false %}
|
| 22 |
+
{%- set ns.last_query_index = index %}
|
| 23 |
+
{%- endif %}
|
| 24 |
+
{%- endfor %}
|
| 25 |
+
{%- for message in messages %}
|
| 26 |
+
{%- if message.content is string %}
|
| 27 |
+
{%- set content = message.content %}
|
| 28 |
+
{%- else %}
|
| 29 |
+
{%- set content = '' %}
|
| 30 |
+
{%- endif %}
|
| 31 |
+
{%- if (message.role == "user") or (message.role == "system" and not loop.first) %}
|
| 32 |
+
{{- '<|im_start|>' + message.role + '\n' + content + '<|im_end|>' + '\n' }}
|
| 33 |
+
{%- elif message.role == "assistant" %}
|
| 34 |
+
{%- set reasoning_content = '' %}
|
| 35 |
+
{%- if message.reasoning_content is string %}
|
| 36 |
+
{%- set reasoning_content = message.reasoning_content %}
|
| 37 |
+
{%- else %}
|
| 38 |
+
{%- if '</think>' in content %}
|
| 39 |
+
{%- set reasoning_content = content.split('</think>')[0].rstrip('\n').split('<think>')[-1].lstrip('\n') %}
|
| 40 |
+
{%- set content = content.split('</think>')[-1].lstrip('\n') %}
|
| 41 |
+
{%- endif %}
|
| 42 |
+
{%- endif %}
|
| 43 |
+
{%- if loop.index0 > ns.last_query_index %}
|
| 44 |
+
{%- if loop.last or (not loop.last and reasoning_content) %}
|
| 45 |
+
{{- '<|im_start|>' + message.role + '\n<think>\n' + reasoning_content.strip('\n') + '\n</think>\n\n' + content.lstrip('\n') }}
|
| 46 |
+
{%- else %}
|
| 47 |
+
{{- '<|im_start|>' + message.role + '\n' + content }}
|
| 48 |
+
{%- endif %}
|
| 49 |
+
{%- else %}
|
| 50 |
+
{{- '<|im_start|>' + message.role + '\n' + content }}
|
| 51 |
+
{%- endif %}
|
| 52 |
+
{%- if message.tool_calls %}
|
| 53 |
+
{%- for tool_call in message.tool_calls %}
|
| 54 |
+
{%- if (loop.first and content) or (not loop.first) %}
|
| 55 |
+
{{- '\n' }}
|
| 56 |
+
{%- endif %}
|
| 57 |
+
{%- if tool_call.function %}
|
| 58 |
+
{%- set tool_call = tool_call.function %}
|
| 59 |
+
{%- endif %}
|
| 60 |
+
{{- '<tool_call>\n{"name": "' }}
|
| 61 |
+
{{- tool_call.name }}
|
| 62 |
+
{{- '", "arguments": ' }}
|
| 63 |
+
{%- if tool_call.arguments is string %}
|
| 64 |
+
{{- tool_call.arguments }}
|
| 65 |
+
{%- else %}
|
| 66 |
+
{{- tool_call.arguments | tojson }}
|
| 67 |
+
{%- endif %}
|
| 68 |
+
{{- '}\n</tool_call>' }}
|
| 69 |
+
{%- endfor %}
|
| 70 |
+
{%- endif %}
|
| 71 |
+
{{- '<|im_end|>\n' }}
|
| 72 |
+
{%- elif message.role == "tool" %}
|
| 73 |
+
{%- if loop.first or (messages[loop.index0 - 1].role != "tool") %}
|
| 74 |
+
{{- '<|im_start|>user' }}
|
| 75 |
+
{%- endif %}
|
| 76 |
+
{{- '\n<tool_response>\n' }}
|
| 77 |
+
{{- content }}
|
| 78 |
+
{{- '\n</tool_response>' }}
|
| 79 |
+
{%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
|
| 80 |
+
{{- '<|im_end|>\n' }}
|
| 81 |
+
{%- endif %}
|
| 82 |
+
{%- endif %}
|
| 83 |
+
{%- endfor %}
|
| 84 |
+
{%- if add_generation_prompt %}
|
| 85 |
+
{{- '<|im_start|>assistant\n' }}
|
| 86 |
+
{%- if enable_thinking is defined and enable_thinking is false %}
|
| 87 |
+
{{- '<think>\n\n</think>\n\n' }}
|
| 88 |
+
{%- endif %}
|
| 89 |
+
{%- endif %}
|
config.json
ADDED
|
@@ -0,0 +1,30 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"architectures": [
|
| 3 |
+
"Qwen3ForCausalLM"
|
| 4 |
+
],
|
| 5 |
+
"attention_bias": false,
|
| 6 |
+
"attention_dropout": 0.0,
|
| 7 |
+
"bos_token_id": 151643,
|
| 8 |
+
"eos_token_id": 151645,
|
| 9 |
+
"head_dim": 128,
|
| 10 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 2048,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 6144,
|
| 14 |
+
"max_position_embeddings": 40960,
|
| 15 |
+
"max_window_layers": 28,
|
| 16 |
+
"model_type": "qwen3",
|
| 17 |
+
"num_attention_heads": 16,
|
| 18 |
+
"num_hidden_layers": 28,
|
| 19 |
+
"num_key_value_heads": 8,
|
| 20 |
+
"rms_norm_eps": 1e-06,
|
| 21 |
+
"rope_scaling": null,
|
| 22 |
+
"rope_theta": 1000000,
|
| 23 |
+
"sliding_window": null,
|
| 24 |
+
"tie_word_embeddings": true,
|
| 25 |
+
"torch_dtype": "bfloat16",
|
| 26 |
+
"transformers_version": "4.52.4",
|
| 27 |
+
"use_cache": false,
|
| 28 |
+
"use_sliding_window": false,
|
| 29 |
+
"vocab_size": 151936
|
| 30 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 151643,
|
| 3 |
+
"do_sample": true,
|
| 4 |
+
"eos_token_id": [
|
| 5 |
+
151645,
|
| 6 |
+
151643
|
| 7 |
+
],
|
| 8 |
+
"pad_token_id": 151643,
|
| 9 |
+
"temperature": 0.6,
|
| 10 |
+
"top_k": 20,
|
| 11 |
+
"top_p": 0.95,
|
| 12 |
+
"transformers_version": "4.52.4"
|
| 13 |
+
}
|
merges.txt
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0c5a1c20c05fbb7b17c703eff948a4f77cf79fcdcf58ea0f80f0fc5ac0510573
|
| 3 |
+
size 3441185608
|
special_tokens_map.json
ADDED
|
@@ -0,0 +1,31 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"additional_special_tokens": [
|
| 3 |
+
"<|im_start|>",
|
| 4 |
+
"<|im_end|>",
|
| 5 |
+
"<|object_ref_start|>",
|
| 6 |
+
"<|object_ref_end|>",
|
| 7 |
+
"<|box_start|>",
|
| 8 |
+
"<|box_end|>",
|
| 9 |
+
"<|quad_start|>",
|
| 10 |
+
"<|quad_end|>",
|
| 11 |
+
"<|vision_start|>",
|
| 12 |
+
"<|vision_end|>",
|
| 13 |
+
"<|vision_pad|>",
|
| 14 |
+
"<|image_pad|>",
|
| 15 |
+
"<|video_pad|>"
|
| 16 |
+
],
|
| 17 |
+
"eos_token": {
|
| 18 |
+
"content": "<|im_end|>",
|
| 19 |
+
"lstrip": false,
|
| 20 |
+
"normalized": false,
|
| 21 |
+
"rstrip": false,
|
| 22 |
+
"single_word": false
|
| 23 |
+
},
|
| 24 |
+
"pad_token": {
|
| 25 |
+
"content": "<|endoftext|>",
|
| 26 |
+
"lstrip": false,
|
| 27 |
+
"normalized": false,
|
| 28 |
+
"rstrip": false,
|
| 29 |
+
"single_word": false
|
| 30 |
+
}
|
| 31 |
+
}
|
tokenizer.json
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:aeb13307a71acd8fe81861d94ad54ab689df773318809eed3cbe794b4492dae4
|
| 3 |
+
size 11422654
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,239 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_bos_token": false,
|
| 3 |
+
"add_prefix_space": false,
|
| 4 |
+
"added_tokens_decoder": {
|
| 5 |
+
"151643": {
|
| 6 |
+
"content": "<|endoftext|>",
|
| 7 |
+
"lstrip": false,
|
| 8 |
+
"normalized": false,
|
| 9 |
+
"rstrip": false,
|
| 10 |
+
"single_word": false,
|
| 11 |
+
"special": true
|
| 12 |
+
},
|
| 13 |
+
"151644": {
|
| 14 |
+
"content": "<|im_start|>",
|
| 15 |
+
"lstrip": false,
|
| 16 |
+
"normalized": false,
|
| 17 |
+
"rstrip": false,
|
| 18 |
+
"single_word": false,
|
| 19 |
+
"special": true
|
| 20 |
+
},
|
| 21 |
+
"151645": {
|
| 22 |
+
"content": "<|im_end|>",
|
| 23 |
+
"lstrip": false,
|
| 24 |
+
"normalized": false,
|
| 25 |
+
"rstrip": false,
|
| 26 |
+
"single_word": false,
|
| 27 |
+
"special": true
|
| 28 |
+
},
|
| 29 |
+
"151646": {
|
| 30 |
+
"content": "<|object_ref_start|>",
|
| 31 |
+
"lstrip": false,
|
| 32 |
+
"normalized": false,
|
| 33 |
+
"rstrip": false,
|
| 34 |
+
"single_word": false,
|
| 35 |
+
"special": true
|
| 36 |
+
},
|
| 37 |
+
"151647": {
|
| 38 |
+
"content": "<|object_ref_end|>",
|
| 39 |
+
"lstrip": false,
|
| 40 |
+
"normalized": false,
|
| 41 |
+
"rstrip": false,
|
| 42 |
+
"single_word": false,
|
| 43 |
+
"special": true
|
| 44 |
+
},
|
| 45 |
+
"151648": {
|
| 46 |
+
"content": "<|box_start|>",
|
| 47 |
+
"lstrip": false,
|
| 48 |
+
"normalized": false,
|
| 49 |
+
"rstrip": false,
|
| 50 |
+
"single_word": false,
|
| 51 |
+
"special": true
|
| 52 |
+
},
|
| 53 |
+
"151649": {
|
| 54 |
+
"content": "<|box_end|>",
|
| 55 |
+
"lstrip": false,
|
| 56 |
+
"normalized": false,
|
| 57 |
+
"rstrip": false,
|
| 58 |
+
"single_word": false,
|
| 59 |
+
"special": true
|
| 60 |
+
},
|
| 61 |
+
"151650": {
|
| 62 |
+
"content": "<|quad_start|>",
|
| 63 |
+
"lstrip": false,
|
| 64 |
+
"normalized": false,
|
| 65 |
+
"rstrip": false,
|
| 66 |
+
"single_word": false,
|
| 67 |
+
"special": true
|
| 68 |
+
},
|
| 69 |
+
"151651": {
|
| 70 |
+
"content": "<|quad_end|>",
|
| 71 |
+
"lstrip": false,
|
| 72 |
+
"normalized": false,
|
| 73 |
+
"rstrip": false,
|
| 74 |
+
"single_word": false,
|
| 75 |
+
"special": true
|
| 76 |
+
},
|
| 77 |
+
"151652": {
|
| 78 |
+
"content": "<|vision_start|>",
|
| 79 |
+
"lstrip": false,
|
| 80 |
+
"normalized": false,
|
| 81 |
+
"rstrip": false,
|
| 82 |
+
"single_word": false,
|
| 83 |
+
"special": true
|
| 84 |
+
},
|
| 85 |
+
"151653": {
|
| 86 |
+
"content": "<|vision_end|>",
|
| 87 |
+
"lstrip": false,
|
| 88 |
+
"normalized": false,
|
| 89 |
+
"rstrip": false,
|
| 90 |
+
"single_word": false,
|
| 91 |
+
"special": true
|
| 92 |
+
},
|
| 93 |
+
"151654": {
|
| 94 |
+
"content": "<|vision_pad|>",
|
| 95 |
+
"lstrip": false,
|
| 96 |
+
"normalized": false,
|
| 97 |
+
"rstrip": false,
|
| 98 |
+
"single_word": false,
|
| 99 |
+
"special": true
|
| 100 |
+
},
|
| 101 |
+
"151655": {
|
| 102 |
+
"content": "<|image_pad|>",
|
| 103 |
+
"lstrip": false,
|
| 104 |
+
"normalized": false,
|
| 105 |
+
"rstrip": false,
|
| 106 |
+
"single_word": false,
|
| 107 |
+
"special": true
|
| 108 |
+
},
|
| 109 |
+
"151656": {
|
| 110 |
+
"content": "<|video_pad|>",
|
| 111 |
+
"lstrip": false,
|
| 112 |
+
"normalized": false,
|
| 113 |
+
"rstrip": false,
|
| 114 |
+
"single_word": false,
|
| 115 |
+
"special": true
|
| 116 |
+
},
|
| 117 |
+
"151657": {
|
| 118 |
+
"content": "<tool_call>",
|
| 119 |
+
"lstrip": false,
|
| 120 |
+
"normalized": false,
|
| 121 |
+
"rstrip": false,
|
| 122 |
+
"single_word": false,
|
| 123 |
+
"special": false
|
| 124 |
+
},
|
| 125 |
+
"151658": {
|
| 126 |
+
"content": "</tool_call>",
|
| 127 |
+
"lstrip": false,
|
| 128 |
+
"normalized": false,
|
| 129 |
+
"rstrip": false,
|
| 130 |
+
"single_word": false,
|
| 131 |
+
"special": false
|
| 132 |
+
},
|
| 133 |
+
"151659": {
|
| 134 |
+
"content": "<|fim_prefix|>",
|
| 135 |
+
"lstrip": false,
|
| 136 |
+
"normalized": false,
|
| 137 |
+
"rstrip": false,
|
| 138 |
+
"single_word": false,
|
| 139 |
+
"special": false
|
| 140 |
+
},
|
| 141 |
+
"151660": {
|
| 142 |
+
"content": "<|fim_middle|>",
|
| 143 |
+
"lstrip": false,
|
| 144 |
+
"normalized": false,
|
| 145 |
+
"rstrip": false,
|
| 146 |
+
"single_word": false,
|
| 147 |
+
"special": false
|
| 148 |
+
},
|
| 149 |
+
"151661": {
|
| 150 |
+
"content": "<|fim_suffix|>",
|
| 151 |
+
"lstrip": false,
|
| 152 |
+
"normalized": false,
|
| 153 |
+
"rstrip": false,
|
| 154 |
+
"single_word": false,
|
| 155 |
+
"special": false
|
| 156 |
+
},
|
| 157 |
+
"151662": {
|
| 158 |
+
"content": "<|fim_pad|>",
|
| 159 |
+
"lstrip": false,
|
| 160 |
+
"normalized": false,
|
| 161 |
+
"rstrip": false,
|
| 162 |
+
"single_word": false,
|
| 163 |
+
"special": false
|
| 164 |
+
},
|
| 165 |
+
"151663": {
|
| 166 |
+
"content": "<|repo_name|>",
|
| 167 |
+
"lstrip": false,
|
| 168 |
+
"normalized": false,
|
| 169 |
+
"rstrip": false,
|
| 170 |
+
"single_word": false,
|
| 171 |
+
"special": false
|
| 172 |
+
},
|
| 173 |
+
"151664": {
|
| 174 |
+
"content": "<|file_sep|>",
|
| 175 |
+
"lstrip": false,
|
| 176 |
+
"normalized": false,
|
| 177 |
+
"rstrip": false,
|
| 178 |
+
"single_word": false,
|
| 179 |
+
"special": false
|
| 180 |
+
},
|
| 181 |
+
"151665": {
|
| 182 |
+
"content": "<tool_response>",
|
| 183 |
+
"lstrip": false,
|
| 184 |
+
"normalized": false,
|
| 185 |
+
"rstrip": false,
|
| 186 |
+
"single_word": false,
|
| 187 |
+
"special": false
|
| 188 |
+
},
|
| 189 |
+
"151666": {
|
| 190 |
+
"content": "</tool_response>",
|
| 191 |
+
"lstrip": false,
|
| 192 |
+
"normalized": false,
|
| 193 |
+
"rstrip": false,
|
| 194 |
+
"single_word": false,
|
| 195 |
+
"special": false
|
| 196 |
+
},
|
| 197 |
+
"151667": {
|
| 198 |
+
"content": "<think>",
|
| 199 |
+
"lstrip": false,
|
| 200 |
+
"normalized": false,
|
| 201 |
+
"rstrip": false,
|
| 202 |
+
"single_word": false,
|
| 203 |
+
"special": false
|
| 204 |
+
},
|
| 205 |
+
"151668": {
|
| 206 |
+
"content": "</think>",
|
| 207 |
+
"lstrip": false,
|
| 208 |
+
"normalized": false,
|
| 209 |
+
"rstrip": false,
|
| 210 |
+
"single_word": false,
|
| 211 |
+
"special": false
|
| 212 |
+
}
|
| 213 |
+
},
|
| 214 |
+
"additional_special_tokens": [
|
| 215 |
+
"<|im_start|>",
|
| 216 |
+
"<|im_end|>",
|
| 217 |
+
"<|object_ref_start|>",
|
| 218 |
+
"<|object_ref_end|>",
|
| 219 |
+
"<|box_start|>",
|
| 220 |
+
"<|box_end|>",
|
| 221 |
+
"<|quad_start|>",
|
| 222 |
+
"<|quad_end|>",
|
| 223 |
+
"<|vision_start|>",
|
| 224 |
+
"<|vision_end|>",
|
| 225 |
+
"<|vision_pad|>",
|
| 226 |
+
"<|image_pad|>",
|
| 227 |
+
"<|video_pad|>"
|
| 228 |
+
],
|
| 229 |
+
"bos_token": null,
|
| 230 |
+
"clean_up_tokenization_spaces": false,
|
| 231 |
+
"eos_token": "<|im_end|>",
|
| 232 |
+
"errors": "replace",
|
| 233 |
+
"extra_special_tokens": {},
|
| 234 |
+
"model_max_length": 131072,
|
| 235 |
+
"pad_token": "<|endoftext|>",
|
| 236 |
+
"split_special_tokens": false,
|
| 237 |
+
"tokenizer_class": "Qwen2Tokenizer",
|
| 238 |
+
"unk_token": null
|
| 239 |
+
}
|
trainer_state.json
ADDED
|
@@ -0,0 +1,1734 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"best_global_step": null,
|
| 3 |
+
"best_metric": null,
|
| 4 |
+
"best_model_checkpoint": null,
|
| 5 |
+
"epoch": 0.0006053268765133172,
|
| 6 |
+
"eval_steps": 500,
|
| 7 |
+
"global_step": 100,
|
| 8 |
+
"is_hyper_param_search": false,
|
| 9 |
+
"is_local_process_zero": true,
|
| 10 |
+
"is_world_process_zero": true,
|
| 11 |
+
"log_history": [
|
| 12 |
+
{
|
| 13 |
+
"clip_ratio/high_max": 0.0,
|
| 14 |
+
"clip_ratio/high_mean": 0.0,
|
| 15 |
+
"clip_ratio/low_mean": 0.0,
|
| 16 |
+
"clip_ratio/low_min": 0.0,
|
| 17 |
+
"clip_ratio/region_mean": 0.0,
|
| 18 |
+
"completions/clipped_ratio": 0.0,
|
| 19 |
+
"completions/max_length": 4095.0,
|
| 20 |
+
"completions/max_terminated_length": 4095.0,
|
| 21 |
+
"completions/mean_length": 1229.75,
|
| 22 |
+
"completions/mean_terminated_length": 1229.75,
|
| 23 |
+
"completions/min_length": 354.0,
|
| 24 |
+
"completions/min_terminated_length": 354.0,
|
| 25 |
+
"epoch": 6.053268765133172e-06,
|
| 26 |
+
"grad_norm": 0.07209322590841183,
|
| 27 |
+
"kl": 0.011932373046875,
|
| 28 |
+
"learning_rate": 0.0,
|
| 29 |
+
"loss": 0.0,
|
| 30 |
+
"num_tokens": 27312.0,
|
| 31 |
+
"reward": 0.0,
|
| 32 |
+
"reward_std": 0.0,
|
| 33 |
+
"rewards/avg_thinking_length_func": 73.625,
|
| 34 |
+
"rewards/correct_answer_reward_func": 0.0,
|
| 35 |
+
"rewards/efficient_thinking_reward_func": 0.45941640084546015,
|
| 36 |
+
"rewards/format_reward_func": 0.987500011920929,
|
| 37 |
+
"rewards/num_xml_reward_func": 1.03125,
|
| 38 |
+
"rewards/tool_execution_reward_func": 0.6875,
|
| 39 |
+
"step": 1
|
| 40 |
+
},
|
| 41 |
+
{
|
| 42 |
+
"clip_ratio/high_max": 0.0,
|
| 43 |
+
"clip_ratio/high_mean": 0.0,
|
| 44 |
+
"clip_ratio/low_mean": 0.0,
|
| 45 |
+
"clip_ratio/low_min": 0.0,
|
| 46 |
+
"clip_ratio/region_mean": 0.0,
|
| 47 |
+
"epoch": 1.2106537530266343e-05,
|
| 48 |
+
"grad_norm": 0.07214504525426423,
|
| 49 |
+
"kl": 0.011932373046875,
|
| 50 |
+
"learning_rate": 1.5000000000000002e-07,
|
| 51 |
+
"loss": 0.0,
|
| 52 |
+
"step": 2
|
| 53 |
+
},
|
| 54 |
+
{
|
| 55 |
+
"clip_ratio/high_max": 0.0,
|
| 56 |
+
"clip_ratio/high_mean": 0.0,
|
| 57 |
+
"clip_ratio/low_mean": 0.0,
|
| 58 |
+
"clip_ratio/low_min": 0.0,
|
| 59 |
+
"clip_ratio/region_mean": 0.0,
|
| 60 |
+
"epoch": 1.8159806295399516e-05,
|
| 61 |
+
"grad_norm": 0.053264765218760744,
|
| 62 |
+
"kl": 0.0120086669921875,
|
| 63 |
+
"learning_rate": 3.0000000000000004e-07,
|
| 64 |
+
"loss": 0.0,
|
| 65 |
+
"step": 3
|
| 66 |
+
},
|
| 67 |
+
{
|
| 68 |
+
"clip_ratio/high_max": 0.0,
|
| 69 |
+
"clip_ratio/high_mean": 0.0,
|
| 70 |
+
"clip_ratio/low_mean": 0.0,
|
| 71 |
+
"clip_ratio/low_min": 0.0,
|
| 72 |
+
"clip_ratio/region_mean": 0.0,
|
| 73 |
+
"epoch": 2.4213075060532686e-05,
|
| 74 |
+
"grad_norm": 0.05122123329371393,
|
| 75 |
+
"kl": 0.01312255859375,
|
| 76 |
+
"learning_rate": 4.5e-07,
|
| 77 |
+
"loss": 0.0,
|
| 78 |
+
"step": 4
|
| 79 |
+
},
|
| 80 |
+
{
|
| 81 |
+
"clip_ratio/high_max": 0.0,
|
| 82 |
+
"clip_ratio/high_mean": 0.0,
|
| 83 |
+
"clip_ratio/low_mean": 0.0,
|
| 84 |
+
"clip_ratio/low_min": 0.0,
|
| 85 |
+
"clip_ratio/region_mean": 0.0,
|
| 86 |
+
"completions/clipped_ratio": 0.0625,
|
| 87 |
+
"completions/max_length": 2499.0,
|
| 88 |
+
"completions/max_terminated_length": 2499.0,
|
| 89 |
+
"completions/mean_length": 1216.875,
|
| 90 |
+
"completions/mean_terminated_length": 1298.0,
|
| 91 |
+
"completions/min_length": 0.0,
|
| 92 |
+
"completions/min_terminated_length": 13.0,
|
| 93 |
+
"epoch": 3.026634382566586e-05,
|
| 94 |
+
"grad_norm": 1.163761459062396,
|
| 95 |
+
"kl": 0.01102447509765625,
|
| 96 |
+
"learning_rate": 6.000000000000001e-07,
|
| 97 |
+
"loss": -0.0148,
|
| 98 |
+
"num_tokens": 58482.0,
|
| 99 |
+
"reward": 0.0473831407725811,
|
| 100 |
+
"reward_std": 0.1895325630903244,
|
| 101 |
+
"rewards/avg_thinking_length_func": 129.625,
|
| 102 |
+
"rewards/correct_answer_reward_func": 0.0625,
|
| 103 |
+
"rewards/efficient_thinking_reward_func": 0.37734568415024294,
|
| 104 |
+
"rewards/format_reward_func": 0.9375,
|
| 105 |
+
"rewards/num_xml_reward_func": 0.9479166269302368,
|
| 106 |
+
"rewards/tool_execution_reward_func": 0.8125,
|
| 107 |
+
"step": 5
|
| 108 |
+
},
|
| 109 |
+
{
|
| 110 |
+
"clip_ratio/high_max": 0.0,
|
| 111 |
+
"clip_ratio/high_mean": 0.0,
|
| 112 |
+
"clip_ratio/low_mean": 0.0,
|
| 113 |
+
"clip_ratio/low_min": 0.0,
|
| 114 |
+
"clip_ratio/region_mean": 0.0,
|
| 115 |
+
"epoch": 3.631961259079903e-05,
|
| 116 |
+
"grad_norm": 1.1267039287882714,
|
| 117 |
+
"kl": 0.0118865966796875,
|
| 118 |
+
"learning_rate": 7.5e-07,
|
| 119 |
+
"loss": -0.0148,
|
| 120 |
+
"step": 6
|
| 121 |
+
},
|
| 122 |
+
{
|
| 123 |
+
"clip_ratio/high_max": 0.0,
|
| 124 |
+
"clip_ratio/high_mean": 0.0,
|
| 125 |
+
"clip_ratio/low_mean": 0.0,
|
| 126 |
+
"clip_ratio/low_min": 0.0,
|
| 127 |
+
"clip_ratio/region_mean": 0.0,
|
| 128 |
+
"epoch": 4.2372881355932206e-05,
|
| 129 |
+
"grad_norm": 1.0659471491371402,
|
| 130 |
+
"kl": 0.0120391845703125,
|
| 131 |
+
"learning_rate": 9e-07,
|
| 132 |
+
"loss": -0.0148,
|
| 133 |
+
"step": 7
|
| 134 |
+
},
|
| 135 |
+
{
|
| 136 |
+
"clip_ratio/high_max": 0.0,
|
| 137 |
+
"clip_ratio/high_mean": 0.0,
|
| 138 |
+
"clip_ratio/low_mean": 0.0,
|
| 139 |
+
"clip_ratio/low_min": 0.0,
|
| 140 |
+
"clip_ratio/region_mean": 0.0,
|
| 141 |
+
"epoch": 4.842615012106537e-05,
|
| 142 |
+
"grad_norm": 1.057686209956945,
|
| 143 |
+
"kl": 0.015655517578125,
|
| 144 |
+
"learning_rate": 1.05e-06,
|
| 145 |
+
"loss": -0.0148,
|
| 146 |
+
"step": 8
|
| 147 |
+
},
|
| 148 |
+
{
|
| 149 |
+
"clip_ratio/high_max": 0.0,
|
| 150 |
+
"clip_ratio/high_mean": 0.0,
|
| 151 |
+
"clip_ratio/low_mean": 0.0,
|
| 152 |
+
"clip_ratio/low_min": 0.0,
|
| 153 |
+
"clip_ratio/region_mean": 0.0,
|
| 154 |
+
"completions/clipped_ratio": 0.0,
|
| 155 |
+
"completions/max_length": 2692.0,
|
| 156 |
+
"completions/max_terminated_length": 2692.0,
|
| 157 |
+
"completions/mean_length": 1255.625,
|
| 158 |
+
"completions/mean_terminated_length": 1255.625,
|
| 159 |
+
"completions/min_length": 224.0,
|
| 160 |
+
"completions/min_terminated_length": 224.0,
|
| 161 |
+
"epoch": 5.4479418886198546e-05,
|
| 162 |
+
"grad_norm": 1.4615219881629928,
|
| 163 |
+
"kl": 0.030914306640625,
|
| 164 |
+
"learning_rate": 1.2000000000000002e-06,
|
| 165 |
+
"loss": -0.0149,
|
| 166 |
+
"num_tokens": 86164.0,
|
| 167 |
+
"reward": 0.16301509737968445,
|
| 168 |
+
"reward_std": 0.29241904616355896,
|
| 169 |
+
"rewards/avg_thinking_length_func": 80.65625,
|
| 170 |
+
"rewards/correct_answer_reward_func": 0.25,
|
| 171 |
+
"rewards/efficient_thinking_reward_func": 0.5424594521082594,
|
| 172 |
+
"rewards/format_reward_func": 1.0,
|
| 173 |
+
"rewards/num_xml_reward_func": 1.1458332538604736,
|
| 174 |
+
"rewards/tool_execution_reward_func": 0.9375,
|
| 175 |
+
"step": 9
|
| 176 |
+
},
|
| 177 |
+
{
|
| 178 |
+
"clip_ratio/high_max": 0.0,
|
| 179 |
+
"clip_ratio/high_mean": 0.0,
|
| 180 |
+
"clip_ratio/low_mean": 0.0,
|
| 181 |
+
"clip_ratio/low_min": 0.0,
|
| 182 |
+
"clip_ratio/region_mean": 0.0,
|
| 183 |
+
"epoch": 6.053268765133172e-05,
|
| 184 |
+
"grad_norm": 1.4154730974531706,
|
| 185 |
+
"kl": 0.0540771484375,
|
| 186 |
+
"learning_rate": 1.35e-06,
|
| 187 |
+
"loss": -0.0149,
|
| 188 |
+
"step": 10
|
| 189 |
+
},
|
| 190 |
+
{
|
| 191 |
+
"clip_ratio/high_max": 0.0,
|
| 192 |
+
"clip_ratio/high_mean": 0.0,
|
| 193 |
+
"clip_ratio/low_mean": 0.0,
|
| 194 |
+
"clip_ratio/low_min": 0.0,
|
| 195 |
+
"clip_ratio/region_mean": 0.0,
|
| 196 |
+
"epoch": 6.658595641646489e-05,
|
| 197 |
+
"grad_norm": 1.330539156016462,
|
| 198 |
+
"kl": 0.0582275390625,
|
| 199 |
+
"learning_rate": 1.5e-06,
|
| 200 |
+
"loss": -0.0148,
|
| 201 |
+
"step": 11
|
| 202 |
+
},
|
| 203 |
+
{
|
| 204 |
+
"clip_ratio/high_max": 0.0,
|
| 205 |
+
"clip_ratio/high_mean": 0.0,
|
| 206 |
+
"clip_ratio/low_mean": 0.0,
|
| 207 |
+
"clip_ratio/low_min": 0.0,
|
| 208 |
+
"clip_ratio/region_mean": 0.0,
|
| 209 |
+
"epoch": 7.263922518159807e-05,
|
| 210 |
+
"grad_norm": 1.152360438708297,
|
| 211 |
+
"kl": 0.1031494140625,
|
| 212 |
+
"learning_rate": 1.65e-06,
|
| 213 |
+
"loss": -0.0147,
|
| 214 |
+
"step": 12
|
| 215 |
+
},
|
| 216 |
+
{
|
| 217 |
+
"clip_ratio/high_max": 0.0,
|
| 218 |
+
"clip_ratio/high_mean": 0.0,
|
| 219 |
+
"clip_ratio/low_mean": 0.0,
|
| 220 |
+
"clip_ratio/low_min": 0.0,
|
| 221 |
+
"clip_ratio/region_mean": 0.0,
|
| 222 |
+
"completions/clipped_ratio": 0.0,
|
| 223 |
+
"completions/max_length": 2962.0,
|
| 224 |
+
"completions/max_terminated_length": 2962.0,
|
| 225 |
+
"completions/mean_length": 1276.4375,
|
| 226 |
+
"completions/mean_terminated_length": 1276.4375,
|
| 227 |
+
"completions/min_length": 13.0,
|
| 228 |
+
"completions/min_terminated_length": 13.0,
|
| 229 |
+
"epoch": 7.869249394673124e-05,
|
| 230 |
+
"grad_norm": 0.6615289048669231,
|
| 231 |
+
"kl": 0.1724853515625,
|
| 232 |
+
"learning_rate": 1.8e-06,
|
| 233 |
+
"loss": 0.0005,
|
| 234 |
+
"num_tokens": 114191.0,
|
| 235 |
+
"reward": 0.0,
|
| 236 |
+
"reward_std": 0.0,
|
| 237 |
+
"rewards/avg_thinking_length_func": 127.6875,
|
| 238 |
+
"rewards/correct_answer_reward_func": 0.0,
|
| 239 |
+
"rewards/efficient_thinking_reward_func": 0.3556471644759768,
|
| 240 |
+
"rewards/format_reward_func": 0.949999988079071,
|
| 241 |
+
"rewards/num_xml_reward_func": 0.875,
|
| 242 |
+
"rewards/tool_execution_reward_func": 0.75,
|
| 243 |
+
"step": 13
|
| 244 |
+
},
|
| 245 |
+
{
|
| 246 |
+
"clip_ratio/high_max": 0.0,
|
| 247 |
+
"clip_ratio/high_mean": 0.0,
|
| 248 |
+
"clip_ratio/low_mean": 0.0,
|
| 249 |
+
"clip_ratio/low_min": 0.0,
|
| 250 |
+
"clip_ratio/region_mean": 0.0,
|
| 251 |
+
"epoch": 8.474576271186441e-05,
|
| 252 |
+
"grad_norm": 0.12138546214366204,
|
| 253 |
+
"kl": 0.147705078125,
|
| 254 |
+
"learning_rate": 1.95e-06,
|
| 255 |
+
"loss": 0.0004,
|
| 256 |
+
"step": 14
|
| 257 |
+
},
|
| 258 |
+
{
|
| 259 |
+
"clip_ratio/high_max": 0.0,
|
| 260 |
+
"clip_ratio/high_mean": 0.0,
|
| 261 |
+
"clip_ratio/low_mean": 0.0,
|
| 262 |
+
"clip_ratio/low_min": 0.0,
|
| 263 |
+
"clip_ratio/region_mean": 0.0,
|
| 264 |
+
"epoch": 9.079903147699757e-05,
|
| 265 |
+
"grad_norm": 0.09332769954040936,
|
| 266 |
+
"kl": 0.1591796875,
|
| 267 |
+
"learning_rate": 2.1e-06,
|
| 268 |
+
"loss": 0.0004,
|
| 269 |
+
"step": 15
|
| 270 |
+
},
|
| 271 |
+
{
|
| 272 |
+
"clip_ratio/high_max": 0.0,
|
| 273 |
+
"clip_ratio/high_mean": 0.0,
|
| 274 |
+
"clip_ratio/low_mean": 0.0,
|
| 275 |
+
"clip_ratio/low_min": 0.0,
|
| 276 |
+
"clip_ratio/region_mean": 0.0,
|
| 277 |
+
"epoch": 9.685230024213075e-05,
|
| 278 |
+
"grad_norm": 0.07716503221980582,
|
| 279 |
+
"kl": 0.162109375,
|
| 280 |
+
"learning_rate": 2.25e-06,
|
| 281 |
+
"loss": 0.0004,
|
| 282 |
+
"step": 16
|
| 283 |
+
},
|
| 284 |
+
{
|
| 285 |
+
"clip_ratio/high_max": 0.0,
|
| 286 |
+
"clip_ratio/high_mean": 0.0,
|
| 287 |
+
"clip_ratio/low_mean": 0.0,
|
| 288 |
+
"clip_ratio/low_min": 0.0,
|
| 289 |
+
"clip_ratio/region_mean": 0.0,
|
| 290 |
+
"completions/clipped_ratio": 0.0625,
|
| 291 |
+
"completions/max_length": 2705.0,
|
| 292 |
+
"completions/max_terminated_length": 2705.0,
|
| 293 |
+
"completions/mean_length": 1008.3125,
|
| 294 |
+
"completions/mean_terminated_length": 1075.5333333333333,
|
| 295 |
+
"completions/min_length": 0.0,
|
| 296 |
+
"completions/min_terminated_length": 302.0,
|
| 297 |
+
"epoch": 0.00010290556900726392,
|
| 298 |
+
"grad_norm": 0.9574173404984538,
|
| 299 |
+
"kl": 0.200439453125,
|
| 300 |
+
"learning_rate": 2.4000000000000003e-06,
|
| 301 |
+
"loss": -0.0132,
|
| 302 |
+
"num_tokens": 142012.0,
|
| 303 |
+
"reward": 0.2393198013305664,
|
| 304 |
+
"reward_std": 0.32144659757614136,
|
| 305 |
+
"rewards/avg_thinking_length_func": 90.72916412353516,
|
| 306 |
+
"rewards/correct_answer_reward_func": 0.375,
|
| 307 |
+
"rewards/efficient_thinking_reward_func": 0.4970963413018923,
|
| 308 |
+
"rewards/format_reward_func": 0.9937499761581421,
|
| 309 |
+
"rewards/num_xml_reward_func": 1.1197916269302368,
|
| 310 |
+
"rewards/tool_execution_reward_func": 0.75,
|
| 311 |
+
"step": 17
|
| 312 |
+
},
|
| 313 |
+
{
|
| 314 |
+
"clip_ratio/high_max": 0.0,
|
| 315 |
+
"clip_ratio/high_mean": 0.0,
|
| 316 |
+
"clip_ratio/low_mean": 0.0,
|
| 317 |
+
"clip_ratio/low_min": 0.0,
|
| 318 |
+
"clip_ratio/region_mean": 0.0,
|
| 319 |
+
"epoch": 0.00010895883777239709,
|
| 320 |
+
"grad_norm": 0.828846190744325,
|
| 321 |
+
"kl": 0.23681640625,
|
| 322 |
+
"learning_rate": 2.55e-06,
|
| 323 |
+
"loss": -0.0131,
|
| 324 |
+
"step": 18
|
| 325 |
+
},
|
| 326 |
+
{
|
| 327 |
+
"clip_ratio/high_max": 0.0,
|
| 328 |
+
"clip_ratio/high_mean": 0.0,
|
| 329 |
+
"clip_ratio/low_mean": 0.0,
|
| 330 |
+
"clip_ratio/low_min": 0.0,
|
| 331 |
+
"clip_ratio/region_mean": 0.0,
|
| 332 |
+
"epoch": 0.00011501210653753027,
|
| 333 |
+
"grad_norm": 1.0886098363962515,
|
| 334 |
+
"kl": 0.34619140625,
|
| 335 |
+
"learning_rate": 2.7e-06,
|
| 336 |
+
"loss": -0.0128,
|
| 337 |
+
"step": 19
|
| 338 |
+
},
|
| 339 |
+
{
|
| 340 |
+
"clip_ratio/high_max": 0.0,
|
| 341 |
+
"clip_ratio/high_mean": 0.0,
|
| 342 |
+
"clip_ratio/low_mean": 0.0,
|
| 343 |
+
"clip_ratio/low_min": 0.0,
|
| 344 |
+
"clip_ratio/region_mean": 0.0,
|
| 345 |
+
"epoch": 0.00012106537530266344,
|
| 346 |
+
"grad_norm": 0.6602921666981004,
|
| 347 |
+
"kl": 0.31396484375,
|
| 348 |
+
"learning_rate": 2.85e-06,
|
| 349 |
+
"loss": -0.0129,
|
| 350 |
+
"step": 20
|
| 351 |
+
},
|
| 352 |
+
{
|
| 353 |
+
"clip_ratio/high_max": 0.0,
|
| 354 |
+
"clip_ratio/high_mean": 0.0,
|
| 355 |
+
"clip_ratio/low_mean": 0.0,
|
| 356 |
+
"clip_ratio/low_min": 0.0,
|
| 357 |
+
"clip_ratio/region_mean": 0.0,
|
| 358 |
+
"completions/clipped_ratio": 0.0625,
|
| 359 |
+
"completions/max_length": 2584.0,
|
| 360 |
+
"completions/max_terminated_length": 2584.0,
|
| 361 |
+
"completions/mean_length": 1127.125,
|
| 362 |
+
"completions/mean_terminated_length": 1202.2666666666667,
|
| 363 |
+
"completions/min_length": 0.0,
|
| 364 |
+
"completions/min_terminated_length": 13.0,
|
| 365 |
+
"epoch": 0.0001271186440677966,
|
| 366 |
+
"grad_norm": 0.29632542403765916,
|
| 367 |
+
"kl": 0.54736328125,
|
| 368 |
+
"learning_rate": 3e-06,
|
| 369 |
+
"loss": 0.0011,
|
| 370 |
+
"num_tokens": 171746.0,
|
| 371 |
+
"reward": 0.0,
|
| 372 |
+
"reward_std": 0.0,
|
| 373 |
+
"rewards/avg_thinking_length_func": 147.97500610351562,
|
| 374 |
+
"rewards/correct_answer_reward_func": 0.0,
|
| 375 |
+
"rewards/efficient_thinking_reward_func": 0.5322939489765828,
|
| 376 |
+
"rewards/format_reward_func": 0.9375,
|
| 377 |
+
"rewards/num_xml_reward_func": 1.0625,
|
| 378 |
+
"rewards/tool_execution_reward_func": 0.800000011920929,
|
| 379 |
+
"step": 21
|
| 380 |
+
},
|
| 381 |
+
{
|
| 382 |
+
"clip_ratio/high_max": 0.0,
|
| 383 |
+
"clip_ratio/high_mean": 0.0,
|
| 384 |
+
"clip_ratio/low_mean": 0.0,
|
| 385 |
+
"clip_ratio/low_min": 0.0,
|
| 386 |
+
"clip_ratio/region_mean": 0.0,
|
| 387 |
+
"epoch": 0.00013317191283292979,
|
| 388 |
+
"grad_norm": 0.1030784592698183,
|
| 389 |
+
"kl": 0.60009765625,
|
| 390 |
+
"learning_rate": 3e-06,
|
| 391 |
+
"loss": 0.0011,
|
| 392 |
+
"step": 22
|
| 393 |
+
},
|
| 394 |
+
{
|
| 395 |
+
"clip_ratio/high_max": 0.0,
|
| 396 |
+
"clip_ratio/high_mean": 0.0,
|
| 397 |
+
"clip_ratio/low_mean": 0.0,
|
| 398 |
+
"clip_ratio/low_min": 0.0,
|
| 399 |
+
"clip_ratio/region_mean": 0.0,
|
| 400 |
+
"epoch": 0.00013922518159806296,
|
| 401 |
+
"grad_norm": 0.06720833154961504,
|
| 402 |
+
"kl": 0.615234375,
|
| 403 |
+
"learning_rate": 3e-06,
|
| 404 |
+
"loss": 0.0011,
|
| 405 |
+
"step": 23
|
| 406 |
+
},
|
| 407 |
+
{
|
| 408 |
+
"clip_ratio/high_max": 0.0,
|
| 409 |
+
"clip_ratio/high_mean": 0.0,
|
| 410 |
+
"clip_ratio/low_mean": 0.0,
|
| 411 |
+
"clip_ratio/low_min": 0.0,
|
| 412 |
+
"clip_ratio/region_mean": 0.0,
|
| 413 |
+
"epoch": 0.00014527845036319613,
|
| 414 |
+
"grad_norm": 0.06007920361540232,
|
| 415 |
+
"kl": 0.5869140625,
|
| 416 |
+
"learning_rate": 3e-06,
|
| 417 |
+
"loss": 0.0011,
|
| 418 |
+
"step": 24
|
| 419 |
+
},
|
| 420 |
+
{
|
| 421 |
+
"clip_ratio/high_max": 0.0,
|
| 422 |
+
"clip_ratio/high_mean": 0.0,
|
| 423 |
+
"clip_ratio/low_mean": 0.0,
|
| 424 |
+
"clip_ratio/low_min": 0.0,
|
| 425 |
+
"clip_ratio/region_mean": 0.0,
|
| 426 |
+
"completions/clipped_ratio": 0.125,
|
| 427 |
+
"completions/max_length": 2784.0,
|
| 428 |
+
"completions/max_terminated_length": 2784.0,
|
| 429 |
+
"completions/mean_length": 1101.5625,
|
| 430 |
+
"completions/mean_terminated_length": 1258.9285714285713,
|
| 431 |
+
"completions/min_length": 0.0,
|
| 432 |
+
"completions/min_terminated_length": 307.0,
|
| 433 |
+
"epoch": 0.0001513317191283293,
|
| 434 |
+
"grad_norm": 89.19553783212066,
|
| 435 |
+
"kl": 9.30859375,
|
| 436 |
+
"learning_rate": 3e-06,
|
| 437 |
+
"loss": 0.0237,
|
| 438 |
+
"num_tokens": 205155.0,
|
| 439 |
+
"reward": -0.10851632058620453,
|
| 440 |
+
"reward_std": 0.300925612449646,
|
| 441 |
+
"rewards/avg_thinking_length_func": 73.3125,
|
| 442 |
+
"rewards/correct_answer_reward_func": 0.25,
|
| 443 |
+
"rewards/efficient_thinking_reward_func": 0.5233806335786315,
|
| 444 |
+
"rewards/format_reward_func": 1.0,
|
| 445 |
+
"rewards/num_xml_reward_func": 1.0416666269302368,
|
| 446 |
+
"rewards/tool_execution_reward_func": 0.9375,
|
| 447 |
+
"step": 25
|
| 448 |
+
},
|
| 449 |
+
{
|
| 450 |
+
"clip_ratio/high_max": 0.0,
|
| 451 |
+
"clip_ratio/high_mean": 0.0,
|
| 452 |
+
"clip_ratio/low_mean": 0.0,
|
| 453 |
+
"clip_ratio/low_min": 0.0,
|
| 454 |
+
"clip_ratio/region_mean": 0.0,
|
| 455 |
+
"epoch": 0.00015738498789346248,
|
| 456 |
+
"grad_norm": 2.0573920189308668,
|
| 457 |
+
"kl": 0.83154296875,
|
| 458 |
+
"learning_rate": 3e-06,
|
| 459 |
+
"loss": 0.0012,
|
| 460 |
+
"step": 26
|
| 461 |
+
},
|
| 462 |
+
{
|
| 463 |
+
"clip_ratio/high_max": 0.0,
|
| 464 |
+
"clip_ratio/high_mean": 0.0,
|
| 465 |
+
"clip_ratio/low_mean": 0.0,
|
| 466 |
+
"clip_ratio/low_min": 0.0,
|
| 467 |
+
"clip_ratio/region_mean": 0.0,
|
| 468 |
+
"epoch": 0.00016343825665859565,
|
| 469 |
+
"grad_norm": 0.8432081368048696,
|
| 470 |
+
"kl": 0.5546875,
|
| 471 |
+
"learning_rate": 3e-06,
|
| 472 |
+
"loss": 0.0004,
|
| 473 |
+
"step": 27
|
| 474 |
+
},
|
| 475 |
+
{
|
| 476 |
+
"clip_ratio/high_max": 0.0,
|
| 477 |
+
"clip_ratio/high_mean": 0.0,
|
| 478 |
+
"clip_ratio/low_mean": 0.0,
|
| 479 |
+
"clip_ratio/low_min": 0.0,
|
| 480 |
+
"clip_ratio/region_mean": 0.0,
|
| 481 |
+
"epoch": 0.00016949152542372882,
|
| 482 |
+
"grad_norm": 1.8587603334809846,
|
| 483 |
+
"kl": 0.546875,
|
| 484 |
+
"learning_rate": 3e-06,
|
| 485 |
+
"loss": 0.0003,
|
| 486 |
+
"step": 28
|
| 487 |
+
},
|
| 488 |
+
{
|
| 489 |
+
"clip_ratio/high_max": 0.0,
|
| 490 |
+
"clip_ratio/high_mean": 0.0,
|
| 491 |
+
"clip_ratio/low_mean": 0.0,
|
| 492 |
+
"clip_ratio/low_min": 0.0,
|
| 493 |
+
"clip_ratio/region_mean": 0.0,
|
| 494 |
+
"completions/clipped_ratio": 0.0,
|
| 495 |
+
"completions/max_length": 2871.0,
|
| 496 |
+
"completions/max_terminated_length": 2871.0,
|
| 497 |
+
"completions/mean_length": 1893.9375,
|
| 498 |
+
"completions/mean_terminated_length": 1893.9375,
|
| 499 |
+
"completions/min_length": 1204.0,
|
| 500 |
+
"completions/min_terminated_length": 1204.0,
|
| 501 |
+
"epoch": 0.000175544794188862,
|
| 502 |
+
"grad_norm": 64.92072887185255,
|
| 503 |
+
"kl": 8.59521484375,
|
| 504 |
+
"learning_rate": 3e-06,
|
| 505 |
+
"loss": 0.0227,
|
| 506 |
+
"num_tokens": 243062.0,
|
| 507 |
+
"reward": 0.03321200609207153,
|
| 508 |
+
"reward_std": 0.13284802436828613,
|
| 509 |
+
"rewards/avg_thinking_length_func": 282.96875,
|
| 510 |
+
"rewards/correct_answer_reward_func": 0.0625,
|
| 511 |
+
"rewards/efficient_thinking_reward_func": 0.4781555664530499,
|
| 512 |
+
"rewards/format_reward_func": 1.0,
|
| 513 |
+
"rewards/num_xml_reward_func": 1.2395832538604736,
|
| 514 |
+
"rewards/tool_execution_reward_func": 1.125,
|
| 515 |
+
"step": 29
|
| 516 |
+
},
|
| 517 |
+
{
|
| 518 |
+
"clip_ratio/high_max": 0.0,
|
| 519 |
+
"clip_ratio/high_mean": 0.0,
|
| 520 |
+
"clip_ratio/low_mean": 0.0,
|
| 521 |
+
"clip_ratio/low_min": 0.0,
|
| 522 |
+
"clip_ratio/region_mean": 0.0,
|
| 523 |
+
"epoch": 0.00018159806295399514,
|
| 524 |
+
"grad_norm": 23076.53534131984,
|
| 525 |
+
"kl": 46.0673828125,
|
| 526 |
+
"learning_rate": 3e-06,
|
| 527 |
+
"loss": 0.3029,
|
| 528 |
+
"step": 30
|
| 529 |
+
},
|
| 530 |
+
{
|
| 531 |
+
"clip_ratio/high_max": 0.0,
|
| 532 |
+
"clip_ratio/high_mean": 0.0,
|
| 533 |
+
"clip_ratio/low_mean": 0.0,
|
| 534 |
+
"clip_ratio/low_min": 0.0,
|
| 535 |
+
"clip_ratio/region_mean": 0.0,
|
| 536 |
+
"epoch": 0.00018765133171912832,
|
| 537 |
+
"grad_norm": 6.2253448587216,
|
| 538 |
+
"kl": 0.97314453125,
|
| 539 |
+
"learning_rate": 3e-06,
|
| 540 |
+
"loss": -0.0007,
|
| 541 |
+
"step": 31
|
| 542 |
+
},
|
| 543 |
+
{
|
| 544 |
+
"clip_ratio/high_max": 0.0,
|
| 545 |
+
"clip_ratio/high_mean": 0.0,
|
| 546 |
+
"clip_ratio/low_mean": 0.0,
|
| 547 |
+
"clip_ratio/low_min": 0.0,
|
| 548 |
+
"clip_ratio/region_mean": 0.0,
|
| 549 |
+
"epoch": 0.0001937046004842615,
|
| 550 |
+
"grad_norm": 3.6768029356119563,
|
| 551 |
+
"kl": 0.62109375,
|
| 552 |
+
"learning_rate": 3e-06,
|
| 553 |
+
"loss": -0.0015,
|
| 554 |
+
"step": 32
|
| 555 |
+
},
|
| 556 |
+
{
|
| 557 |
+
"clip_ratio/high_max": 0.0,
|
| 558 |
+
"clip_ratio/high_mean": 0.0,
|
| 559 |
+
"clip_ratio/low_mean": 0.0,
|
| 560 |
+
"clip_ratio/low_min": 0.0,
|
| 561 |
+
"clip_ratio/region_mean": 0.0,
|
| 562 |
+
"completions/clipped_ratio": 0.0625,
|
| 563 |
+
"completions/max_length": 2817.0,
|
| 564 |
+
"completions/max_terminated_length": 2817.0,
|
| 565 |
+
"completions/mean_length": 1192.4375,
|
| 566 |
+
"completions/mean_terminated_length": 1271.9333333333334,
|
| 567 |
+
"completions/min_length": 0.0,
|
| 568 |
+
"completions/min_terminated_length": 323.0,
|
| 569 |
+
"epoch": 0.00019975786924939466,
|
| 570 |
+
"grad_norm": 3.7035403316845934,
|
| 571 |
+
"kl": 1.03271484375,
|
| 572 |
+
"learning_rate": 3e-06,
|
| 573 |
+
"loss": -0.0039,
|
| 574 |
+
"num_tokens": 273829.0,
|
| 575 |
+
"reward": 0.09496757388114929,
|
| 576 |
+
"reward_std": 0.41386595368385315,
|
| 577 |
+
"rewards/avg_thinking_length_func": 69.1875,
|
| 578 |
+
"rewards/correct_answer_reward_func": 0.4375,
|
| 579 |
+
"rewards/efficient_thinking_reward_func": 0.4750871147644119,
|
| 580 |
+
"rewards/format_reward_func": 1.0,
|
| 581 |
+
"rewards/num_xml_reward_func": 0.9479166269302368,
|
| 582 |
+
"rewards/tool_execution_reward_func": 0.9375,
|
| 583 |
+
"step": 33
|
| 584 |
+
},
|
| 585 |
+
{
|
| 586 |
+
"clip_ratio/high_max": 0.0,
|
| 587 |
+
"clip_ratio/high_mean": 0.0,
|
| 588 |
+
"clip_ratio/low_mean": 0.0,
|
| 589 |
+
"clip_ratio/low_min": 0.0,
|
| 590 |
+
"clip_ratio/region_mean": 0.0,
|
| 591 |
+
"epoch": 0.00020581113801452784,
|
| 592 |
+
"grad_norm": 1.6964420133725846,
|
| 593 |
+
"kl": 0.6962890625,
|
| 594 |
+
"learning_rate": 3e-06,
|
| 595 |
+
"loss": -0.0055,
|
| 596 |
+
"step": 34
|
| 597 |
+
},
|
| 598 |
+
{
|
| 599 |
+
"clip_ratio/high_max": 0.0,
|
| 600 |
+
"clip_ratio/high_mean": 0.0,
|
| 601 |
+
"clip_ratio/low_mean": 0.0,
|
| 602 |
+
"clip_ratio/low_min": 0.0,
|
| 603 |
+
"clip_ratio/region_mean": 0.0,
|
| 604 |
+
"epoch": 0.000211864406779661,
|
| 605 |
+
"grad_norm": 1.2613994802405568,
|
| 606 |
+
"kl": 0.7099609375,
|
| 607 |
+
"learning_rate": 3e-06,
|
| 608 |
+
"loss": -0.0054,
|
| 609 |
+
"step": 35
|
| 610 |
+
},
|
| 611 |
+
{
|
| 612 |
+
"clip_ratio/high_max": 0.0,
|
| 613 |
+
"clip_ratio/high_mean": 0.0,
|
| 614 |
+
"clip_ratio/low_mean": 0.0,
|
| 615 |
+
"clip_ratio/low_min": 0.0,
|
| 616 |
+
"clip_ratio/region_mean": 0.0,
|
| 617 |
+
"epoch": 0.00021791767554479418,
|
| 618 |
+
"grad_norm": 1.4140209746173513,
|
| 619 |
+
"kl": 0.7392578125,
|
| 620 |
+
"learning_rate": 3e-06,
|
| 621 |
+
"loss": -0.0051,
|
| 622 |
+
"step": 36
|
| 623 |
+
},
|
| 624 |
+
{
|
| 625 |
+
"clip_ratio/high_max": 0.0,
|
| 626 |
+
"clip_ratio/high_mean": 0.0,
|
| 627 |
+
"clip_ratio/low_mean": 0.0,
|
| 628 |
+
"clip_ratio/low_min": 0.0,
|
| 629 |
+
"clip_ratio/region_mean": 0.0,
|
| 630 |
+
"completions/clipped_ratio": 0.0625,
|
| 631 |
+
"completions/max_length": 3204.0,
|
| 632 |
+
"completions/max_terminated_length": 3204.0,
|
| 633 |
+
"completions/mean_length": 1571.1875,
|
| 634 |
+
"completions/mean_terminated_length": 1675.9333333333334,
|
| 635 |
+
"completions/min_length": 0.0,
|
| 636 |
+
"completions/min_terminated_length": 511.0,
|
| 637 |
+
"epoch": 0.00022397094430992736,
|
| 638 |
+
"grad_norm": 53596994705358.31,
|
| 639 |
+
"kl": 3710857740297.5625,
|
| 640 |
+
"learning_rate": 3e-06,
|
| 641 |
+
"loss": 10334783488.0,
|
| 642 |
+
"num_tokens": 310668.0,
|
| 643 |
+
"reward": -0.043196823447942734,
|
| 644 |
+
"reward_std": 0.17278729379177094,
|
| 645 |
+
"rewards/avg_thinking_length_func": 162.5625,
|
| 646 |
+
"rewards/correct_answer_reward_func": 0.0625,
|
| 647 |
+
"rewards/efficient_thinking_reward_func": 0.3250826003494903,
|
| 648 |
+
"rewards/format_reward_func": 0.987500011920929,
|
| 649 |
+
"rewards/num_xml_reward_func": 0.875,
|
| 650 |
+
"rewards/tool_execution_reward_func": 0.9375,
|
| 651 |
+
"step": 37
|
| 652 |
+
},
|
| 653 |
+
{
|
| 654 |
+
"clip_ratio/high_max": 0.0,
|
| 655 |
+
"clip_ratio/high_mean": 0.0,
|
| 656 |
+
"clip_ratio/low_mean": 0.0,
|
| 657 |
+
"clip_ratio/low_min": 0.0,
|
| 658 |
+
"clip_ratio/region_mean": 0.0,
|
| 659 |
+
"epoch": 0.00023002421307506053,
|
| 660 |
+
"grad_norm": 84771335.30009712,
|
| 661 |
+
"kl": 2411058.75,
|
| 662 |
+
"learning_rate": 3e-06,
|
| 663 |
+
"loss": 6696.0161,
|
| 664 |
+
"step": 38
|
| 665 |
+
},
|
| 666 |
+
{
|
| 667 |
+
"clip_ratio/high_max": 0.0,
|
| 668 |
+
"clip_ratio/high_mean": 0.0,
|
| 669 |
+
"clip_ratio/low_mean": 0.0,
|
| 670 |
+
"clip_ratio/low_min": 0.0,
|
| 671 |
+
"clip_ratio/region_mean": 0.0,
|
| 672 |
+
"epoch": 0.0002360774818401937,
|
| 673 |
+
"grad_norm": 3632.7359114591427,
|
| 674 |
+
"kl": 133.154296875,
|
| 675 |
+
"learning_rate": 3e-06,
|
| 676 |
+
"loss": 0.82,
|
| 677 |
+
"step": 39
|
| 678 |
+
},
|
| 679 |
+
{
|
| 680 |
+
"clip_ratio/high_max": 0.0,
|
| 681 |
+
"clip_ratio/high_mean": 0.0,
|
| 682 |
+
"clip_ratio/low_mean": 0.0,
|
| 683 |
+
"clip_ratio/low_min": 0.0,
|
| 684 |
+
"clip_ratio/region_mean": 0.0,
|
| 685 |
+
"epoch": 0.00024213075060532688,
|
| 686 |
+
"grad_norm": 42.64315100889087,
|
| 687 |
+
"kl": 3.703125,
|
| 688 |
+
"learning_rate": 3e-06,
|
| 689 |
+
"loss": 0.0345,
|
| 690 |
+
"step": 40
|
| 691 |
+
},
|
| 692 |
+
{
|
| 693 |
+
"clip_ratio/high_max": 0.0,
|
| 694 |
+
"clip_ratio/high_mean": 0.0,
|
| 695 |
+
"clip_ratio/low_mean": 0.0,
|
| 696 |
+
"clip_ratio/low_min": 0.0,
|
| 697 |
+
"clip_ratio/region_mean": 0.0,
|
| 698 |
+
"completions/clipped_ratio": 0.0625,
|
| 699 |
+
"completions/max_length": 2535.0,
|
| 700 |
+
"completions/max_terminated_length": 2535.0,
|
| 701 |
+
"completions/mean_length": 1004.0625,
|
| 702 |
+
"completions/mean_terminated_length": 1071.0,
|
| 703 |
+
"completions/min_length": 0.0,
|
| 704 |
+
"completions/min_terminated_length": 277.0,
|
| 705 |
+
"epoch": 0.00024818401937046,
|
| 706 |
+
"grad_norm": 0.3830041877799282,
|
| 707 |
+
"kl": 0.59814453125,
|
| 708 |
+
"learning_rate": 3e-06,
|
| 709 |
+
"loss": -0.0042,
|
| 710 |
+
"num_tokens": 338421.0,
|
| 711 |
+
"reward": 0.0777856633067131,
|
| 712 |
+
"reward_std": 0.305923730134964,
|
| 713 |
+
"rewards/avg_thinking_length_func": 89.21875,
|
| 714 |
+
"rewards/correct_answer_reward_func": 0.3125,
|
| 715 |
+
"rewards/efficient_thinking_reward_func": 0.49057807654986807,
|
| 716 |
+
"rewards/format_reward_func": 1.0,
|
| 717 |
+
"rewards/num_xml_reward_func": 1.0104166269302368,
|
| 718 |
+
"rewards/tool_execution_reward_func": 0.8125,
|
| 719 |
+
"step": 41
|
| 720 |
+
},
|
| 721 |
+
{
|
| 722 |
+
"clip_ratio/high_max": 0.0,
|
| 723 |
+
"clip_ratio/high_mean": 0.0,
|
| 724 |
+
"clip_ratio/low_mean": 0.0,
|
| 725 |
+
"clip_ratio/low_min": 0.0,
|
| 726 |
+
"clip_ratio/region_mean": 0.0,
|
| 727 |
+
"epoch": 0.0002542372881355932,
|
| 728 |
+
"grad_norm": 0.26752404691559056,
|
| 729 |
+
"kl": 0.572265625,
|
| 730 |
+
"learning_rate": 3e-06,
|
| 731 |
+
"loss": -0.0043,
|
| 732 |
+
"step": 42
|
| 733 |
+
},
|
| 734 |
+
{
|
| 735 |
+
"clip_ratio/high_max": 0.0,
|
| 736 |
+
"clip_ratio/high_mean": 0.0,
|
| 737 |
+
"clip_ratio/low_mean": 0.0,
|
| 738 |
+
"clip_ratio/low_min": 0.0,
|
| 739 |
+
"clip_ratio/region_mean": 0.0,
|
| 740 |
+
"epoch": 0.00026029055690072637,
|
| 741 |
+
"grad_norm": 0.24396627644436175,
|
| 742 |
+
"kl": 0.59033203125,
|
| 743 |
+
"learning_rate": 3e-06,
|
| 744 |
+
"loss": -0.0043,
|
| 745 |
+
"step": 43
|
| 746 |
+
},
|
| 747 |
+
{
|
| 748 |
+
"clip_ratio/high_max": 0.0,
|
| 749 |
+
"clip_ratio/high_mean": 0.0,
|
| 750 |
+
"clip_ratio/low_mean": 0.0,
|
| 751 |
+
"clip_ratio/low_min": 0.0,
|
| 752 |
+
"clip_ratio/region_mean": 0.0,
|
| 753 |
+
"epoch": 0.00026634382566585957,
|
| 754 |
+
"grad_norm": 0.2219905360560463,
|
| 755 |
+
"kl": 0.5869140625,
|
| 756 |
+
"learning_rate": 3e-06,
|
| 757 |
+
"loss": -0.0043,
|
| 758 |
+
"step": 44
|
| 759 |
+
},
|
| 760 |
+
{
|
| 761 |
+
"clip_ratio/high_max": 0.0,
|
| 762 |
+
"clip_ratio/high_mean": 0.0,
|
| 763 |
+
"clip_ratio/low_mean": 0.0,
|
| 764 |
+
"clip_ratio/low_min": 0.0,
|
| 765 |
+
"clip_ratio/region_mean": 0.0,
|
| 766 |
+
"completions/clipped_ratio": 0.0,
|
| 767 |
+
"completions/max_length": 2556.0,
|
| 768 |
+
"completions/max_terminated_length": 2556.0,
|
| 769 |
+
"completions/mean_length": 1469.9375,
|
| 770 |
+
"completions/mean_terminated_length": 1469.9375,
|
| 771 |
+
"completions/min_length": 806.0,
|
| 772 |
+
"completions/min_terminated_length": 806.0,
|
| 773 |
+
"epoch": 0.0002723970944309927,
|
| 774 |
+
"grad_norm": 0.3505513257767094,
|
| 775 |
+
"kl": 0.515625,
|
| 776 |
+
"learning_rate": 3e-06,
|
| 777 |
+
"loss": 0.0017,
|
| 778 |
+
"num_tokens": 369544.0,
|
| 779 |
+
"reward": 0.0,
|
| 780 |
+
"reward_std": 0.0,
|
| 781 |
+
"rewards/avg_thinking_length_func": 145.875,
|
| 782 |
+
"rewards/correct_answer_reward_func": 0.0,
|
| 783 |
+
"rewards/efficient_thinking_reward_func": 0.3200705292271429,
|
| 784 |
+
"rewards/format_reward_func": 0.9937499761581421,
|
| 785 |
+
"rewards/num_xml_reward_func": 0.875,
|
| 786 |
+
"rewards/tool_execution_reward_func": 0.875,
|
| 787 |
+
"step": 45
|
| 788 |
+
},
|
| 789 |
+
{
|
| 790 |
+
"clip_ratio/high_max": 0.0,
|
| 791 |
+
"clip_ratio/high_mean": 0.0,
|
| 792 |
+
"clip_ratio/low_mean": 0.0,
|
| 793 |
+
"clip_ratio/low_min": 0.0,
|
| 794 |
+
"clip_ratio/region_mean": 0.0,
|
| 795 |
+
"epoch": 0.0002784503631961259,
|
| 796 |
+
"grad_norm": 0.1358254536364525,
|
| 797 |
+
"kl": 0.48291015625,
|
| 798 |
+
"learning_rate": 3e-06,
|
| 799 |
+
"loss": 0.0016,
|
| 800 |
+
"step": 46
|
| 801 |
+
},
|
| 802 |
+
{
|
| 803 |
+
"clip_ratio/high_max": 0.0,
|
| 804 |
+
"clip_ratio/high_mean": 0.0,
|
| 805 |
+
"clip_ratio/low_mean": 0.0,
|
| 806 |
+
"clip_ratio/low_min": 0.0,
|
| 807 |
+
"clip_ratio/region_mean": 0.0,
|
| 808 |
+
"epoch": 0.00028450363196125906,
|
| 809 |
+
"grad_norm": 0.04953296360188178,
|
| 810 |
+
"kl": 0.46240234375,
|
| 811 |
+
"learning_rate": 3e-06,
|
| 812 |
+
"loss": 0.0016,
|
| 813 |
+
"step": 47
|
| 814 |
+
},
|
| 815 |
+
{
|
| 816 |
+
"clip_ratio/high_max": 0.0,
|
| 817 |
+
"clip_ratio/high_mean": 0.0,
|
| 818 |
+
"clip_ratio/low_mean": 0.0,
|
| 819 |
+
"clip_ratio/low_min": 0.0,
|
| 820 |
+
"clip_ratio/region_mean": 0.0,
|
| 821 |
+
"epoch": 0.00029055690072639226,
|
| 822 |
+
"grad_norm": 0.04556158513173327,
|
| 823 |
+
"kl": 0.458984375,
|
| 824 |
+
"learning_rate": 3e-06,
|
| 825 |
+
"loss": 0.0015,
|
| 826 |
+
"step": 48
|
| 827 |
+
},
|
| 828 |
+
{
|
| 829 |
+
"clip_ratio/high_max": 0.0,
|
| 830 |
+
"clip_ratio/high_mean": 0.0,
|
| 831 |
+
"clip_ratio/low_mean": 0.0,
|
| 832 |
+
"clip_ratio/low_min": 0.0,
|
| 833 |
+
"clip_ratio/region_mean": 0.0,
|
| 834 |
+
"completions/clipped_ratio": 0.0,
|
| 835 |
+
"completions/max_length": 1418.0,
|
| 836 |
+
"completions/max_terminated_length": 1418.0,
|
| 837 |
+
"completions/mean_length": 1150.0625,
|
| 838 |
+
"completions/mean_terminated_length": 1150.0625,
|
| 839 |
+
"completions/min_length": 281.0,
|
| 840 |
+
"completions/min_terminated_length": 281.0,
|
| 841 |
+
"epoch": 0.0002966101694915254,
|
| 842 |
+
"grad_norm": 24.523298976288224,
|
| 843 |
+
"kl": 4.013671875,
|
| 844 |
+
"learning_rate": 3e-06,
|
| 845 |
+
"loss": 0.0174,
|
| 846 |
+
"num_tokens": 395537.0,
|
| 847 |
+
"reward": 0.09051951766014099,
|
| 848 |
+
"reward_std": 0.40777578949928284,
|
| 849 |
+
"rewards/avg_thinking_length_func": 90.625,
|
| 850 |
+
"rewards/correct_answer_reward_func": 0.4375,
|
| 851 |
+
"rewards/efficient_thinking_reward_func": 0.6560945998667584,
|
| 852 |
+
"rewards/format_reward_func": 1.0,
|
| 853 |
+
"rewards/num_xml_reward_func": 1.1875,
|
| 854 |
+
"rewards/tool_execution_reward_func": 0.9375,
|
| 855 |
+
"step": 49
|
| 856 |
+
},
|
| 857 |
+
{
|
| 858 |
+
"clip_ratio/high_max": 0.0,
|
| 859 |
+
"clip_ratio/high_mean": 0.0,
|
| 860 |
+
"clip_ratio/low_mean": 0.0,
|
| 861 |
+
"clip_ratio/low_min": 0.0,
|
| 862 |
+
"clip_ratio/region_mean": 0.0,
|
| 863 |
+
"epoch": 0.0003026634382566586,
|
| 864 |
+
"grad_norm": 1.1088188455815937,
|
| 865 |
+
"kl": 1.0966796875,
|
| 866 |
+
"learning_rate": 3e-06,
|
| 867 |
+
"loss": 0.0085,
|
| 868 |
+
"step": 50
|
| 869 |
+
},
|
| 870 |
+
{
|
| 871 |
+
"clip_ratio/high_max": 0.0,
|
| 872 |
+
"clip_ratio/high_mean": 0.0,
|
| 873 |
+
"clip_ratio/low_mean": 0.0,
|
| 874 |
+
"clip_ratio/low_min": 0.0,
|
| 875 |
+
"clip_ratio/region_mean": 0.0,
|
| 876 |
+
"epoch": 0.00030871670702179176,
|
| 877 |
+
"grad_norm": 0.8246471166001736,
|
| 878 |
+
"kl": 0.998046875,
|
| 879 |
+
"learning_rate": 3e-06,
|
| 880 |
+
"loss": 0.0081,
|
| 881 |
+
"step": 51
|
| 882 |
+
},
|
| 883 |
+
{
|
| 884 |
+
"clip_ratio/high_max": 0.0,
|
| 885 |
+
"clip_ratio/high_mean": 0.0,
|
| 886 |
+
"clip_ratio/low_mean": 0.0,
|
| 887 |
+
"clip_ratio/low_min": 0.0,
|
| 888 |
+
"clip_ratio/region_mean": 0.0,
|
| 889 |
+
"epoch": 0.00031476997578692496,
|
| 890 |
+
"grad_norm": 0.7444859388030179,
|
| 891 |
+
"kl": 0.927734375,
|
| 892 |
+
"learning_rate": 3e-06,
|
| 893 |
+
"loss": 0.0079,
|
| 894 |
+
"step": 52
|
| 895 |
+
},
|
| 896 |
+
{
|
| 897 |
+
"clip_ratio/high_max": 0.0,
|
| 898 |
+
"clip_ratio/high_mean": 0.0,
|
| 899 |
+
"clip_ratio/low_mean": 0.0,
|
| 900 |
+
"clip_ratio/low_min": 0.0,
|
| 901 |
+
"clip_ratio/region_mean": 0.0,
|
| 902 |
+
"completions/clipped_ratio": 0.0,
|
| 903 |
+
"completions/max_length": 3603.0,
|
| 904 |
+
"completions/max_terminated_length": 3603.0,
|
| 905 |
+
"completions/mean_length": 1481.75,
|
| 906 |
+
"completions/mean_terminated_length": 1481.75,
|
| 907 |
+
"completions/min_length": 938.0,
|
| 908 |
+
"completions/min_terminated_length": 938.0,
|
| 909 |
+
"epoch": 0.0003208232445520581,
|
| 910 |
+
"grad_norm": 0.6560467591027961,
|
| 911 |
+
"kl": 0.72265625,
|
| 912 |
+
"learning_rate": 3e-06,
|
| 913 |
+
"loss": 0.0026,
|
| 914 |
+
"num_tokens": 426849.0,
|
| 915 |
+
"reward": 0.0,
|
| 916 |
+
"reward_std": 0.0,
|
| 917 |
+
"rewards/avg_thinking_length_func": 110.40625,
|
| 918 |
+
"rewards/correct_answer_reward_func": 0.0,
|
| 919 |
+
"rewards/efficient_thinking_reward_func": 0.5525294297041425,
|
| 920 |
+
"rewards/format_reward_func": 0.9697916507720947,
|
| 921 |
+
"rewards/num_xml_reward_func": 0.9583333730697632,
|
| 922 |
+
"rewards/tool_execution_reward_func": 0.875,
|
| 923 |
+
"step": 53
|
| 924 |
+
},
|
| 925 |
+
{
|
| 926 |
+
"clip_ratio/high_max": 0.0,
|
| 927 |
+
"clip_ratio/high_mean": 0.0,
|
| 928 |
+
"clip_ratio/low_mean": 0.0,
|
| 929 |
+
"clip_ratio/low_min": 0.0,
|
| 930 |
+
"clip_ratio/region_mean": 0.0,
|
| 931 |
+
"epoch": 0.0003268765133171913,
|
| 932 |
+
"grad_norm": 0.9054808017375268,
|
| 933 |
+
"kl": 0.7802734375,
|
| 934 |
+
"learning_rate": 3e-06,
|
| 935 |
+
"loss": 0.0027,
|
| 936 |
+
"step": 54
|
| 937 |
+
},
|
| 938 |
+
{
|
| 939 |
+
"clip_ratio/high_max": 0.0,
|
| 940 |
+
"clip_ratio/high_mean": 0.0,
|
| 941 |
+
"clip_ratio/low_mean": 0.0,
|
| 942 |
+
"clip_ratio/low_min": 0.0,
|
| 943 |
+
"clip_ratio/region_mean": 0.0,
|
| 944 |
+
"epoch": 0.00033292978208232445,
|
| 945 |
+
"grad_norm": 0.2664167272731428,
|
| 946 |
+
"kl": 0.6806640625,
|
| 947 |
+
"learning_rate": 3e-06,
|
| 948 |
+
"loss": 0.0024,
|
| 949 |
+
"step": 55
|
| 950 |
+
},
|
| 951 |
+
{
|
| 952 |
+
"clip_ratio/high_max": 0.0,
|
| 953 |
+
"clip_ratio/high_mean": 0.0,
|
| 954 |
+
"clip_ratio/low_mean": 0.0,
|
| 955 |
+
"clip_ratio/low_min": 0.0,
|
| 956 |
+
"clip_ratio/region_mean": 0.0,
|
| 957 |
+
"epoch": 0.00033898305084745765,
|
| 958 |
+
"grad_norm": 0.06723837861451398,
|
| 959 |
+
"kl": 0.6181640625,
|
| 960 |
+
"learning_rate": 3e-06,
|
| 961 |
+
"loss": 0.0022,
|
| 962 |
+
"step": 56
|
| 963 |
+
},
|
| 964 |
+
{
|
| 965 |
+
"clip_ratio/high_max": 0.0,
|
| 966 |
+
"clip_ratio/high_mean": 0.0,
|
| 967 |
+
"clip_ratio/low_mean": 0.0,
|
| 968 |
+
"clip_ratio/low_min": 0.0,
|
| 969 |
+
"clip_ratio/region_mean": 0.0,
|
| 970 |
+
"completions/clipped_ratio": 0.0,
|
| 971 |
+
"completions/max_length": 3084.0,
|
| 972 |
+
"completions/max_terminated_length": 3084.0,
|
| 973 |
+
"completions/mean_length": 1275.125,
|
| 974 |
+
"completions/mean_terminated_length": 1275.125,
|
| 975 |
+
"completions/min_length": 239.0,
|
| 976 |
+
"completions/min_terminated_length": 239.0,
|
| 977 |
+
"epoch": 0.0003450363196125908,
|
| 978 |
+
"grad_norm": 18149.022369023623,
|
| 979 |
+
"kl": 436.88671875,
|
| 980 |
+
"learning_rate": 3e-06,
|
| 981 |
+
"loss": 2.4098,
|
| 982 |
+
"num_tokens": 454843.0,
|
| 983 |
+
"reward": 0.06402035802602768,
|
| 984 |
+
"reward_std": 0.3948812782764435,
|
| 985 |
+
"rewards/avg_thinking_length_func": 72.0,
|
| 986 |
+
"rewards/correct_answer_reward_func": 0.375,
|
| 987 |
+
"rewards/efficient_thinking_reward_func": 0.5193150989217454,
|
| 988 |
+
"rewards/format_reward_func": 0.9791666865348816,
|
| 989 |
+
"rewards/num_xml_reward_func": 1.046875,
|
| 990 |
+
"rewards/tool_execution_reward_func": 0.875,
|
| 991 |
+
"step": 57
|
| 992 |
+
},
|
| 993 |
+
{
|
| 994 |
+
"clip_ratio/high_max": 0.0,
|
| 995 |
+
"clip_ratio/high_mean": 0.0,
|
| 996 |
+
"clip_ratio/low_mean": 0.0,
|
| 997 |
+
"clip_ratio/low_min": 0.0,
|
| 998 |
+
"clip_ratio/region_mean": 0.0,
|
| 999 |
+
"epoch": 0.000351089588377724,
|
| 1000 |
+
"grad_norm": 130.24455511173358,
|
| 1001 |
+
"kl": 25.609375,
|
| 1002 |
+
"learning_rate": 3e-06,
|
| 1003 |
+
"loss": 0.0696,
|
| 1004 |
+
"step": 58
|
| 1005 |
+
},
|
| 1006 |
+
{
|
| 1007 |
+
"clip_ratio/high_max": 0.0,
|
| 1008 |
+
"clip_ratio/high_mean": 0.0,
|
| 1009 |
+
"clip_ratio/low_mean": 0.0,
|
| 1010 |
+
"clip_ratio/low_min": 0.0,
|
| 1011 |
+
"clip_ratio/region_mean": 0.0,
|
| 1012 |
+
"epoch": 0.00035714285714285714,
|
| 1013 |
+
"grad_norm": 6.994583625280491,
|
| 1014 |
+
"kl": 3.1640625,
|
| 1015 |
+
"learning_rate": 3e-06,
|
| 1016 |
+
"loss": -0.0064,
|
| 1017 |
+
"step": 59
|
| 1018 |
+
},
|
| 1019 |
+
{
|
| 1020 |
+
"clip_ratio/high_max": 0.0,
|
| 1021 |
+
"clip_ratio/high_mean": 0.0,
|
| 1022 |
+
"clip_ratio/low_mean": 0.0,
|
| 1023 |
+
"clip_ratio/low_min": 0.0,
|
| 1024 |
+
"clip_ratio/region_mean": 0.0,
|
| 1025 |
+
"epoch": 0.0003631961259079903,
|
| 1026 |
+
"grad_norm": 1.0257205408467853,
|
| 1027 |
+
"kl": 1.396484375,
|
| 1028 |
+
"learning_rate": 3e-06,
|
| 1029 |
+
"loss": -0.0124,
|
| 1030 |
+
"step": 60
|
| 1031 |
+
},
|
| 1032 |
+
{
|
| 1033 |
+
"clip_ratio/high_max": 0.0,
|
| 1034 |
+
"clip_ratio/high_mean": 0.0,
|
| 1035 |
+
"clip_ratio/low_mean": 0.0,
|
| 1036 |
+
"clip_ratio/low_min": 0.0,
|
| 1037 |
+
"clip_ratio/region_mean": 0.0,
|
| 1038 |
+
"completions/clipped_ratio": 0.0625,
|
| 1039 |
+
"completions/max_length": 3353.0,
|
| 1040 |
+
"completions/max_terminated_length": 3353.0,
|
| 1041 |
+
"completions/mean_length": 1478.4375,
|
| 1042 |
+
"completions/mean_terminated_length": 1577.0,
|
| 1043 |
+
"completions/min_length": 0.0,
|
| 1044 |
+
"completions/min_terminated_length": 13.0,
|
| 1045 |
+
"epoch": 0.0003692493946731235,
|
| 1046 |
+
"grad_norm": 0.39241037722147315,
|
| 1047 |
+
"kl": 0.7880859375,
|
| 1048 |
+
"learning_rate": 3e-06,
|
| 1049 |
+
"loss": -0.0248,
|
| 1050 |
+
"num_tokens": 490198.0,
|
| 1051 |
+
"reward": 0.004050910472869873,
|
| 1052 |
+
"reward_std": 0.2644372284412384,
|
| 1053 |
+
"rewards/avg_thinking_length_func": 157.6770782470703,
|
| 1054 |
+
"rewards/correct_answer_reward_func": 0.125,
|
| 1055 |
+
"rewards/efficient_thinking_reward_func": 0.5676122758117079,
|
| 1056 |
+
"rewards/format_reward_func": 0.9004167318344116,
|
| 1057 |
+
"rewards/num_xml_reward_func": 0.96875,
|
| 1058 |
+
"rewards/tool_execution_reward_func": 0.9375,
|
| 1059 |
+
"step": 61
|
| 1060 |
+
},
|
| 1061 |
+
{
|
| 1062 |
+
"clip_ratio/high_max": 0.0,
|
| 1063 |
+
"clip_ratio/high_mean": 0.0,
|
| 1064 |
+
"clip_ratio/low_mean": 0.0,
|
| 1065 |
+
"clip_ratio/low_min": 0.0,
|
| 1066 |
+
"clip_ratio/region_mean": 0.0,
|
| 1067 |
+
"epoch": 0.00037530266343825664,
|
| 1068 |
+
"grad_norm": 0.33115302407577757,
|
| 1069 |
+
"kl": 0.78515625,
|
| 1070 |
+
"learning_rate": 3e-06,
|
| 1071 |
+
"loss": -0.0248,
|
| 1072 |
+
"step": 62
|
| 1073 |
+
},
|
| 1074 |
+
{
|
| 1075 |
+
"clip_ratio/high_max": 0.0,
|
| 1076 |
+
"clip_ratio/high_mean": 0.0,
|
| 1077 |
+
"clip_ratio/low_mean": 0.0,
|
| 1078 |
+
"clip_ratio/low_min": 0.0,
|
| 1079 |
+
"clip_ratio/region_mean": 0.0,
|
| 1080 |
+
"epoch": 0.00038135593220338984,
|
| 1081 |
+
"grad_norm": 0.25505560853744225,
|
| 1082 |
+
"kl": 0.798828125,
|
| 1083 |
+
"learning_rate": 3e-06,
|
| 1084 |
+
"loss": -0.0248,
|
| 1085 |
+
"step": 63
|
| 1086 |
+
},
|
| 1087 |
+
{
|
| 1088 |
+
"clip_ratio/high_max": 0.0,
|
| 1089 |
+
"clip_ratio/high_mean": 0.0,
|
| 1090 |
+
"clip_ratio/low_mean": 0.0,
|
| 1091 |
+
"clip_ratio/low_min": 0.0,
|
| 1092 |
+
"clip_ratio/region_mean": 0.0,
|
| 1093 |
+
"epoch": 0.000387409200968523,
|
| 1094 |
+
"grad_norm": 0.2102809253373098,
|
| 1095 |
+
"kl": 0.828125,
|
| 1096 |
+
"learning_rate": 3e-06,
|
| 1097 |
+
"loss": -0.0248,
|
| 1098 |
+
"step": 64
|
| 1099 |
+
},
|
| 1100 |
+
{
|
| 1101 |
+
"clip_ratio/high_max": 0.0,
|
| 1102 |
+
"clip_ratio/high_mean": 0.0,
|
| 1103 |
+
"clip_ratio/low_mean": 0.0,
|
| 1104 |
+
"clip_ratio/low_min": 0.0,
|
| 1105 |
+
"clip_ratio/region_mean": 0.0,
|
| 1106 |
+
"completions/clipped_ratio": 0.0625,
|
| 1107 |
+
"completions/max_length": 2543.0,
|
| 1108 |
+
"completions/max_terminated_length": 2543.0,
|
| 1109 |
+
"completions/mean_length": 1185.3125,
|
| 1110 |
+
"completions/mean_terminated_length": 1264.3333333333333,
|
| 1111 |
+
"completions/min_length": 0.0,
|
| 1112 |
+
"completions/min_terminated_length": 213.0,
|
| 1113 |
+
"epoch": 0.0003934624697336562,
|
| 1114 |
+
"grad_norm": 0.8562023922194224,
|
| 1115 |
+
"kl": 1.2021484375,
|
| 1116 |
+
"learning_rate": 3e-06,
|
| 1117 |
+
"loss": 0.0037,
|
| 1118 |
+
"num_tokens": 520851.0,
|
| 1119 |
+
"reward": 0.0788969099521637,
|
| 1120 |
+
"reward_std": 0.2156054973602295,
|
| 1121 |
+
"rewards/avg_thinking_length_func": 114.40625,
|
| 1122 |
+
"rewards/correct_answer_reward_func": 0.125,
|
| 1123 |
+
"rewards/efficient_thinking_reward_func": 0.7406251293101025,
|
| 1124 |
+
"rewards/format_reward_func": 0.9466667175292969,
|
| 1125 |
+
"rewards/num_xml_reward_func": 1.0520833730697632,
|
| 1126 |
+
"rewards/tool_execution_reward_func": 0.875,
|
| 1127 |
+
"step": 65
|
| 1128 |
+
},
|
| 1129 |
+
{
|
| 1130 |
+
"clip_ratio/high_max": 0.0,
|
| 1131 |
+
"clip_ratio/high_mean": 0.0,
|
| 1132 |
+
"clip_ratio/low_mean": 0.0,
|
| 1133 |
+
"clip_ratio/low_min": 0.0,
|
| 1134 |
+
"clip_ratio/region_mean": 0.0,
|
| 1135 |
+
"epoch": 0.00039951573849878933,
|
| 1136 |
+
"grad_norm": 0.2046311200261111,
|
| 1137 |
+
"kl": 1.0009765625,
|
| 1138 |
+
"learning_rate": 3e-06,
|
| 1139 |
+
"loss": 0.0031,
|
| 1140 |
+
"step": 66
|
| 1141 |
+
},
|
| 1142 |
+
{
|
| 1143 |
+
"clip_ratio/high_max": 0.0,
|
| 1144 |
+
"clip_ratio/high_mean": 0.0,
|
| 1145 |
+
"clip_ratio/low_mean": 0.0,
|
| 1146 |
+
"clip_ratio/low_min": 0.0,
|
| 1147 |
+
"clip_ratio/region_mean": 0.0,
|
| 1148 |
+
"epoch": 0.00040556900726392253,
|
| 1149 |
+
"grad_norm": 0.1947907245304449,
|
| 1150 |
+
"kl": 0.9833984375,
|
| 1151 |
+
"learning_rate": 3e-06,
|
| 1152 |
+
"loss": 0.003,
|
| 1153 |
+
"step": 67
|
| 1154 |
+
},
|
| 1155 |
+
{
|
| 1156 |
+
"clip_ratio/high_max": 0.0,
|
| 1157 |
+
"clip_ratio/high_mean": 0.0,
|
| 1158 |
+
"clip_ratio/low_mean": 0.0,
|
| 1159 |
+
"clip_ratio/low_min": 0.0,
|
| 1160 |
+
"clip_ratio/region_mean": 0.0,
|
| 1161 |
+
"epoch": 0.0004116222760290557,
|
| 1162 |
+
"grad_norm": 0.19275059130452188,
|
| 1163 |
+
"kl": 0.9736328125,
|
| 1164 |
+
"learning_rate": 3e-06,
|
| 1165 |
+
"loss": 0.003,
|
| 1166 |
+
"step": 68
|
| 1167 |
+
},
|
| 1168 |
+
{
|
| 1169 |
+
"clip_ratio/high_max": 0.0,
|
| 1170 |
+
"clip_ratio/high_mean": 0.0,
|
| 1171 |
+
"clip_ratio/low_mean": 0.0,
|
| 1172 |
+
"clip_ratio/low_min": 0.0,
|
| 1173 |
+
"clip_ratio/region_mean": 0.0,
|
| 1174 |
+
"completions/clipped_ratio": 0.0,
|
| 1175 |
+
"completions/max_length": 3362.0,
|
| 1176 |
+
"completions/max_terminated_length": 3362.0,
|
| 1177 |
+
"completions/mean_length": 1588.1875,
|
| 1178 |
+
"completions/mean_terminated_length": 1588.1875,
|
| 1179 |
+
"completions/min_length": 252.0,
|
| 1180 |
+
"completions/min_terminated_length": 252.0,
|
| 1181 |
+
"epoch": 0.0004176755447941889,
|
| 1182 |
+
"grad_norm": 0.8366730039548783,
|
| 1183 |
+
"kl": 0.8349609375,
|
| 1184 |
+
"learning_rate": 3e-06,
|
| 1185 |
+
"loss": 0.0219,
|
| 1186 |
+
"num_tokens": 553866.0,
|
| 1187 |
+
"reward": -0.043196823447942734,
|
| 1188 |
+
"reward_std": 0.17278729379177094,
|
| 1189 |
+
"rewards/avg_thinking_length_func": 197.28125,
|
| 1190 |
+
"rewards/correct_answer_reward_func": 0.0625,
|
| 1191 |
+
"rewards/efficient_thinking_reward_func": 0.6236665630221623,
|
| 1192 |
+
"rewards/format_reward_func": 0.984375,
|
| 1193 |
+
"rewards/num_xml_reward_func": 1.375,
|
| 1194 |
+
"rewards/tool_execution_reward_func": 1.0625,
|
| 1195 |
+
"step": 69
|
| 1196 |
+
},
|
| 1197 |
+
{
|
| 1198 |
+
"clip_ratio/high_max": 0.0,
|
| 1199 |
+
"clip_ratio/high_mean": 0.0,
|
| 1200 |
+
"clip_ratio/low_mean": 0.0,
|
| 1201 |
+
"clip_ratio/low_min": 0.0,
|
| 1202 |
+
"clip_ratio/region_mean": 0.0,
|
| 1203 |
+
"epoch": 0.000423728813559322,
|
| 1204 |
+
"grad_norm": 0.9243760422793489,
|
| 1205 |
+
"kl": 0.8310546875,
|
| 1206 |
+
"learning_rate": 3e-06,
|
| 1207 |
+
"loss": 0.0219,
|
| 1208 |
+
"step": 70
|
| 1209 |
+
},
|
| 1210 |
+
{
|
| 1211 |
+
"clip_ratio/high_max": 0.0,
|
| 1212 |
+
"clip_ratio/high_mean": 0.0,
|
| 1213 |
+
"clip_ratio/low_mean": 0.0,
|
| 1214 |
+
"clip_ratio/low_min": 0.0,
|
| 1215 |
+
"clip_ratio/region_mean": 0.0,
|
| 1216 |
+
"epoch": 0.0004297820823244552,
|
| 1217 |
+
"grad_norm": 2.5683071328894296,
|
| 1218 |
+
"kl": 0.947265625,
|
| 1219 |
+
"learning_rate": 3e-06,
|
| 1220 |
+
"loss": 0.0226,
|
| 1221 |
+
"step": 71
|
| 1222 |
+
},
|
| 1223 |
+
{
|
| 1224 |
+
"clip_ratio/high_max": 0.0,
|
| 1225 |
+
"clip_ratio/high_mean": 0.0,
|
| 1226 |
+
"clip_ratio/low_mean": 0.0,
|
| 1227 |
+
"clip_ratio/low_min": 0.0,
|
| 1228 |
+
"clip_ratio/region_mean": 0.0,
|
| 1229 |
+
"epoch": 0.00043583535108958837,
|
| 1230 |
+
"grad_norm": 8580.97667222067,
|
| 1231 |
+
"kl": 167.6015625,
|
| 1232 |
+
"learning_rate": 3e-06,
|
| 1233 |
+
"loss": 1.3871,
|
| 1234 |
+
"step": 72
|
| 1235 |
+
},
|
| 1236 |
+
{
|
| 1237 |
+
"clip_ratio/high_max": 0.0,
|
| 1238 |
+
"clip_ratio/high_mean": 0.0,
|
| 1239 |
+
"clip_ratio/low_mean": 0.0,
|
| 1240 |
+
"clip_ratio/low_min": 0.0,
|
| 1241 |
+
"clip_ratio/region_mean": 0.0,
|
| 1242 |
+
"completions/clipped_ratio": 0.0,
|
| 1243 |
+
"completions/max_length": 2605.0,
|
| 1244 |
+
"completions/max_terminated_length": 2605.0,
|
| 1245 |
+
"completions/mean_length": 1085.4375,
|
| 1246 |
+
"completions/mean_terminated_length": 1085.4375,
|
| 1247 |
+
"completions/min_length": 243.0,
|
| 1248 |
+
"completions/min_terminated_length": 243.0,
|
| 1249 |
+
"epoch": 0.00044188861985472157,
|
| 1250 |
+
"grad_norm": 2.698763686067543,
|
| 1251 |
+
"kl": 1.203125,
|
| 1252 |
+
"learning_rate": 3e-06,
|
| 1253 |
+
"loss": 0.0159,
|
| 1254 |
+
"num_tokens": 578825.0,
|
| 1255 |
+
"reward": 0.13226403295993805,
|
| 1256 |
+
"reward_std": 0.32802054286003113,
|
| 1257 |
+
"rewards/avg_thinking_length_func": 51.78125,
|
| 1258 |
+
"rewards/correct_answer_reward_func": 0.4375,
|
| 1259 |
+
"rewards/efficient_thinking_reward_func": 0.4221610379909578,
|
| 1260 |
+
"rewards/format_reward_func": 1.0,
|
| 1261 |
+
"rewards/num_xml_reward_func": 1.1979166269302368,
|
| 1262 |
+
"rewards/tool_execution_reward_func": 0.8125,
|
| 1263 |
+
"step": 73
|
| 1264 |
+
},
|
| 1265 |
+
{
|
| 1266 |
+
"clip_ratio/high_max": 0.0,
|
| 1267 |
+
"clip_ratio/high_mean": 0.0,
|
| 1268 |
+
"clip_ratio/low_mean": 0.0,
|
| 1269 |
+
"clip_ratio/low_min": 0.0,
|
| 1270 |
+
"clip_ratio/region_mean": 0.0,
|
| 1271 |
+
"epoch": 0.0004479418886198547,
|
| 1272 |
+
"grad_norm": 0.4428441474501145,
|
| 1273 |
+
"kl": 0.9765625,
|
| 1274 |
+
"learning_rate": 3e-06,
|
| 1275 |
+
"loss": 0.0152,
|
| 1276 |
+
"step": 74
|
| 1277 |
+
},
|
| 1278 |
+
{
|
| 1279 |
+
"clip_ratio/high_max": 0.0,
|
| 1280 |
+
"clip_ratio/high_mean": 0.0,
|
| 1281 |
+
"clip_ratio/low_mean": 0.0,
|
| 1282 |
+
"clip_ratio/low_min": 0.0,
|
| 1283 |
+
"clip_ratio/region_mean": 0.0,
|
| 1284 |
+
"epoch": 0.0004539951573849879,
|
| 1285 |
+
"grad_norm": 0.4303895229040926,
|
| 1286 |
+
"kl": 0.9765625,
|
| 1287 |
+
"learning_rate": 3e-06,
|
| 1288 |
+
"loss": 0.0152,
|
| 1289 |
+
"step": 75
|
| 1290 |
+
},
|
| 1291 |
+
{
|
| 1292 |
+
"clip_ratio/high_max": 0.0,
|
| 1293 |
+
"clip_ratio/high_mean": 0.0,
|
| 1294 |
+
"clip_ratio/low_mean": 0.0,
|
| 1295 |
+
"clip_ratio/low_min": 0.0,
|
| 1296 |
+
"clip_ratio/region_mean": 0.0,
|
| 1297 |
+
"epoch": 0.00046004842615012106,
|
| 1298 |
+
"grad_norm": 14.390819598577476,
|
| 1299 |
+
"kl": 2.109375,
|
| 1300 |
+
"learning_rate": 3e-06,
|
| 1301 |
+
"loss": 0.0187,
|
| 1302 |
+
"step": 76
|
| 1303 |
+
},
|
| 1304 |
+
{
|
| 1305 |
+
"clip_ratio/high_max": 0.0,
|
| 1306 |
+
"clip_ratio/high_mean": 0.0,
|
| 1307 |
+
"clip_ratio/low_mean": 0.0,
|
| 1308 |
+
"clip_ratio/low_min": 0.0,
|
| 1309 |
+
"clip_ratio/region_mean": 0.0,
|
| 1310 |
+
"completions/clipped_ratio": 0.0,
|
| 1311 |
+
"completions/max_length": 2957.0,
|
| 1312 |
+
"completions/max_terminated_length": 2957.0,
|
| 1313 |
+
"completions/mean_length": 1399.0625,
|
| 1314 |
+
"completions/mean_terminated_length": 1399.0625,
|
| 1315 |
+
"completions/min_length": 568.0,
|
| 1316 |
+
"completions/min_terminated_length": 568.0,
|
| 1317 |
+
"epoch": 0.00046610169491525426,
|
| 1318 |
+
"grad_norm": 4189.702981431435,
|
| 1319 |
+
"kl": 92.814453125,
|
| 1320 |
+
"learning_rate": 3e-06,
|
| 1321 |
+
"loss": 0.6434,
|
| 1322 |
+
"num_tokens": 608814.0,
|
| 1323 |
+
"reward": 0.044644102454185486,
|
| 1324 |
+
"reward_std": 0.17857640981674194,
|
| 1325 |
+
"rewards/avg_thinking_length_func": 100.25,
|
| 1326 |
+
"rewards/correct_answer_reward_func": 0.0625,
|
| 1327 |
+
"rewards/efficient_thinking_reward_func": 0.4614093326791774,
|
| 1328 |
+
"rewards/format_reward_func": 0.981249988079071,
|
| 1329 |
+
"rewards/num_xml_reward_func": 1.2708332538604736,
|
| 1330 |
+
"rewards/tool_execution_reward_func": 0.875,
|
| 1331 |
+
"step": 77
|
| 1332 |
+
},
|
| 1333 |
+
{
|
| 1334 |
+
"clip_ratio/high_max": 0.0,
|
| 1335 |
+
"clip_ratio/high_mean": 0.0,
|
| 1336 |
+
"clip_ratio/low_mean": 0.0,
|
| 1337 |
+
"clip_ratio/low_min": 0.0,
|
| 1338 |
+
"clip_ratio/region_mean": 0.0,
|
| 1339 |
+
"epoch": 0.0004721549636803874,
|
| 1340 |
+
"grad_norm": 99.00548055580344,
|
| 1341 |
+
"kl": 14.109375,
|
| 1342 |
+
"learning_rate": 3e-06,
|
| 1343 |
+
"loss": 0.0301,
|
| 1344 |
+
"step": 78
|
| 1345 |
+
},
|
| 1346 |
+
{
|
| 1347 |
+
"clip_ratio/high_max": 0.0,
|
| 1348 |
+
"clip_ratio/high_mean": 0.0,
|
| 1349 |
+
"clip_ratio/low_mean": 0.0,
|
| 1350 |
+
"clip_ratio/low_min": 0.0,
|
| 1351 |
+
"clip_ratio/region_mean": 0.0,
|
| 1352 |
+
"epoch": 0.00047820823244552055,
|
| 1353 |
+
"grad_norm": 5.827885456645575,
|
| 1354 |
+
"kl": 2.306640625,
|
| 1355 |
+
"learning_rate": 3e-06,
|
| 1356 |
+
"loss": -0.005,
|
| 1357 |
+
"step": 79
|
| 1358 |
+
},
|
| 1359 |
+
{
|
| 1360 |
+
"clip_ratio/high_max": 0.0,
|
| 1361 |
+
"clip_ratio/high_mean": 0.0,
|
| 1362 |
+
"clip_ratio/low_mean": 0.0,
|
| 1363 |
+
"clip_ratio/low_min": 0.0,
|
| 1364 |
+
"clip_ratio/region_mean": 0.0,
|
| 1365 |
+
"epoch": 0.00048426150121065375,
|
| 1366 |
+
"grad_norm": 1.393289989818607,
|
| 1367 |
+
"kl": 1.1181640625,
|
| 1368 |
+
"learning_rate": 3e-06,
|
| 1369 |
+
"loss": -0.0072,
|
| 1370 |
+
"step": 80
|
| 1371 |
+
},
|
| 1372 |
+
{
|
| 1373 |
+
"clip_ratio/high_max": 0.0,
|
| 1374 |
+
"clip_ratio/high_mean": 0.0,
|
| 1375 |
+
"clip_ratio/low_mean": 0.0,
|
| 1376 |
+
"clip_ratio/low_min": 0.0,
|
| 1377 |
+
"clip_ratio/region_mean": 0.0,
|
| 1378 |
+
"completions/clipped_ratio": 0.0625,
|
| 1379 |
+
"completions/max_length": 2912.0,
|
| 1380 |
+
"completions/max_terminated_length": 2912.0,
|
| 1381 |
+
"completions/mean_length": 1180.6875,
|
| 1382 |
+
"completions/mean_terminated_length": 1259.4,
|
| 1383 |
+
"completions/min_length": 0.0,
|
| 1384 |
+
"completions/min_terminated_length": 318.0,
|
| 1385 |
+
"epoch": 0.000490314769975787,
|
| 1386 |
+
"grad_norm": 15.423239339093648,
|
| 1387 |
+
"kl": 3.791015625,
|
| 1388 |
+
"learning_rate": 3e-06,
|
| 1389 |
+
"loss": 0.0403,
|
| 1390 |
+
"num_tokens": 639393.0,
|
| 1391 |
+
"reward": 0.12507787346839905,
|
| 1392 |
+
"reward_std": 0.42275041341781616,
|
| 1393 |
+
"rewards/avg_thinking_length_func": 77.41666412353516,
|
| 1394 |
+
"rewards/correct_answer_reward_func": 0.5,
|
| 1395 |
+
"rewards/efficient_thinking_reward_func": 0.5368496096788951,
|
| 1396 |
+
"rewards/format_reward_func": 0.9825000166893005,
|
| 1397 |
+
"rewards/num_xml_reward_func": 1.09375,
|
| 1398 |
+
"rewards/tool_execution_reward_func": 0.9375,
|
| 1399 |
+
"step": 81
|
| 1400 |
+
},
|
| 1401 |
+
{
|
| 1402 |
+
"clip_ratio/high_max": 0.0,
|
| 1403 |
+
"clip_ratio/high_mean": 0.0,
|
| 1404 |
+
"clip_ratio/low_mean": 0.0,
|
| 1405 |
+
"clip_ratio/low_min": 0.0,
|
| 1406 |
+
"clip_ratio/region_mean": 0.0,
|
| 1407 |
+
"epoch": 0.00049636803874092,
|
| 1408 |
+
"grad_norm": 2.262950442352818,
|
| 1409 |
+
"kl": 1.6640625,
|
| 1410 |
+
"learning_rate": 3e-06,
|
| 1411 |
+
"loss": 0.032,
|
| 1412 |
+
"step": 82
|
| 1413 |
+
},
|
| 1414 |
+
{
|
| 1415 |
+
"clip_ratio/high_max": 0.0,
|
| 1416 |
+
"clip_ratio/high_mean": 0.0,
|
| 1417 |
+
"clip_ratio/low_mean": 0.0,
|
| 1418 |
+
"clip_ratio/low_min": 0.0,
|
| 1419 |
+
"clip_ratio/region_mean": 0.0,
|
| 1420 |
+
"epoch": 0.0005024213075060532,
|
| 1421 |
+
"grad_norm": 1.1984612569756452,
|
| 1422 |
+
"kl": 1.18359375,
|
| 1423 |
+
"learning_rate": 3e-06,
|
| 1424 |
+
"loss": 0.0306,
|
| 1425 |
+
"step": 83
|
| 1426 |
+
},
|
| 1427 |
+
{
|
| 1428 |
+
"clip_ratio/high_max": 0.0,
|
| 1429 |
+
"clip_ratio/high_mean": 0.0,
|
| 1430 |
+
"clip_ratio/low_mean": 0.0,
|
| 1431 |
+
"clip_ratio/low_min": 0.0,
|
| 1432 |
+
"clip_ratio/region_mean": 0.0,
|
| 1433 |
+
"epoch": 0.0005084745762711864,
|
| 1434 |
+
"grad_norm": 2.42271341300058,
|
| 1435 |
+
"kl": 1.287109375,
|
| 1436 |
+
"learning_rate": 3e-06,
|
| 1437 |
+
"loss": 0.0313,
|
| 1438 |
+
"step": 84
|
| 1439 |
+
},
|
| 1440 |
+
{
|
| 1441 |
+
"clip_ratio/high_max": 0.0,
|
| 1442 |
+
"clip_ratio/high_mean": 0.0,
|
| 1443 |
+
"clip_ratio/low_mean": 0.0,
|
| 1444 |
+
"clip_ratio/low_min": 0.0,
|
| 1445 |
+
"clip_ratio/region_mean": 0.0,
|
| 1446 |
+
"completions/clipped_ratio": 0.125,
|
| 1447 |
+
"completions/max_length": 2703.0,
|
| 1448 |
+
"completions/max_terminated_length": 2703.0,
|
| 1449 |
+
"completions/mean_length": 1237.0,
|
| 1450 |
+
"completions/mean_terminated_length": 1413.7142857142858,
|
| 1451 |
+
"completions/min_length": 0.0,
|
| 1452 |
+
"completions/min_terminated_length": 744.0,
|
| 1453 |
+
"epoch": 0.0005145278450363196,
|
| 1454 |
+
"grad_norm": 0.18115764428552125,
|
| 1455 |
+
"kl": 0.751953125,
|
| 1456 |
+
"learning_rate": 3e-06,
|
| 1457 |
+
"loss": -0.0039,
|
| 1458 |
+
"num_tokens": 674981.0,
|
| 1459 |
+
"reward": 0.03594072163105011,
|
| 1460 |
+
"reward_std": 0.14376287162303925,
|
| 1461 |
+
"rewards/avg_thinking_length_func": 100.3125,
|
| 1462 |
+
"rewards/correct_answer_reward_func": 0.0625,
|
| 1463 |
+
"rewards/efficient_thinking_reward_func": 0.6207628181651339,
|
| 1464 |
+
"rewards/format_reward_func": 0.96875,
|
| 1465 |
+
"rewards/num_xml_reward_func": 1.3270833492279053,
|
| 1466 |
+
"rewards/tool_execution_reward_func": 0.9375,
|
| 1467 |
+
"step": 85
|
| 1468 |
+
},
|
| 1469 |
+
{
|
| 1470 |
+
"clip_ratio/high_max": 0.0,
|
| 1471 |
+
"clip_ratio/high_mean": 0.0,
|
| 1472 |
+
"clip_ratio/low_mean": 0.0,
|
| 1473 |
+
"clip_ratio/low_min": 0.0,
|
| 1474 |
+
"clip_ratio/region_mean": 0.0,
|
| 1475 |
+
"epoch": 0.0005205811138014527,
|
| 1476 |
+
"grad_norm": 0.1563801639696718,
|
| 1477 |
+
"kl": 0.7373046875,
|
| 1478 |
+
"learning_rate": 3e-06,
|
| 1479 |
+
"loss": -0.004,
|
| 1480 |
+
"step": 86
|
| 1481 |
+
},
|
| 1482 |
+
{
|
| 1483 |
+
"clip_ratio/high_max": 0.0,
|
| 1484 |
+
"clip_ratio/high_mean": 0.0,
|
| 1485 |
+
"clip_ratio/low_mean": 0.0,
|
| 1486 |
+
"clip_ratio/low_min": 0.0,
|
| 1487 |
+
"clip_ratio/region_mean": 0.0,
|
| 1488 |
+
"epoch": 0.0005266343825665859,
|
| 1489 |
+
"grad_norm": 0.1436249859111408,
|
| 1490 |
+
"kl": 0.73046875,
|
| 1491 |
+
"learning_rate": 3e-06,
|
| 1492 |
+
"loss": -0.004,
|
| 1493 |
+
"step": 87
|
| 1494 |
+
},
|
| 1495 |
+
{
|
| 1496 |
+
"clip_ratio/high_max": 0.0,
|
| 1497 |
+
"clip_ratio/high_mean": 0.0,
|
| 1498 |
+
"clip_ratio/low_mean": 0.0,
|
| 1499 |
+
"clip_ratio/low_min": 0.0,
|
| 1500 |
+
"clip_ratio/region_mean": 0.0,
|
| 1501 |
+
"epoch": 0.0005326876513317191,
|
| 1502 |
+
"grad_norm": 0.1274894764193074,
|
| 1503 |
+
"kl": 0.724609375,
|
| 1504 |
+
"learning_rate": 3e-06,
|
| 1505 |
+
"loss": -0.004,
|
| 1506 |
+
"step": 88
|
| 1507 |
+
},
|
| 1508 |
+
{
|
| 1509 |
+
"clip_ratio/high_max": 0.0,
|
| 1510 |
+
"clip_ratio/high_mean": 0.0,
|
| 1511 |
+
"clip_ratio/low_mean": 0.0,
|
| 1512 |
+
"clip_ratio/low_min": 0.0,
|
| 1513 |
+
"clip_ratio/region_mean": 0.0,
|
| 1514 |
+
"completions/clipped_ratio": 0.0,
|
| 1515 |
+
"completions/max_length": 2842.0,
|
| 1516 |
+
"completions/max_terminated_length": 2842.0,
|
| 1517 |
+
"completions/mean_length": 1241.375,
|
| 1518 |
+
"completions/mean_terminated_length": 1241.375,
|
| 1519 |
+
"completions/min_length": 230.0,
|
| 1520 |
+
"completions/min_terminated_length": 230.0,
|
| 1521 |
+
"epoch": 0.0005387409200968523,
|
| 1522 |
+
"grad_norm": 5.853699967839861,
|
| 1523 |
+
"kl": 1.76171875,
|
| 1524 |
+
"learning_rate": 3e-06,
|
| 1525 |
+
"loss": 0.0202,
|
| 1526 |
+
"num_tokens": 702435.0,
|
| 1527 |
+
"reward": 0.18937985599040985,
|
| 1528 |
+
"reward_std": 0.26354894042015076,
|
| 1529 |
+
"rewards/avg_thinking_length_func": 74.375,
|
| 1530 |
+
"rewards/correct_answer_reward_func": 0.375,
|
| 1531 |
+
"rewards/efficient_thinking_reward_func": 0.485505122320828,
|
| 1532 |
+
"rewards/format_reward_func": 0.9916666746139526,
|
| 1533 |
+
"rewards/num_xml_reward_func": 1.3125,
|
| 1534 |
+
"rewards/tool_execution_reward_func": 0.8125,
|
| 1535 |
+
"step": 89
|
| 1536 |
+
},
|
| 1537 |
+
{
|
| 1538 |
+
"clip_ratio/high_max": 0.0,
|
| 1539 |
+
"clip_ratio/high_mean": 0.0,
|
| 1540 |
+
"clip_ratio/low_mean": 0.0,
|
| 1541 |
+
"clip_ratio/low_min": 0.0,
|
| 1542 |
+
"clip_ratio/region_mean": 0.0,
|
| 1543 |
+
"epoch": 0.0005447941888619854,
|
| 1544 |
+
"grad_norm": 1.095172664109551,
|
| 1545 |
+
"kl": 1.482421875,
|
| 1546 |
+
"learning_rate": 3e-06,
|
| 1547 |
+
"loss": 0.0181,
|
| 1548 |
+
"step": 90
|
| 1549 |
+
},
|
| 1550 |
+
{
|
| 1551 |
+
"clip_ratio/high_max": 0.0,
|
| 1552 |
+
"clip_ratio/high_mean": 0.0,
|
| 1553 |
+
"clip_ratio/low_mean": 0.0,
|
| 1554 |
+
"clip_ratio/low_min": 0.0,
|
| 1555 |
+
"clip_ratio/region_mean": 0.0,
|
| 1556 |
+
"epoch": 0.0005508474576271186,
|
| 1557 |
+
"grad_norm": 0.42798964827778446,
|
| 1558 |
+
"kl": 1.3359375,
|
| 1559 |
+
"learning_rate": 3e-06,
|
| 1560 |
+
"loss": 0.0174,
|
| 1561 |
+
"step": 91
|
| 1562 |
+
},
|
| 1563 |
+
{
|
| 1564 |
+
"clip_ratio/high_max": 0.0,
|
| 1565 |
+
"clip_ratio/high_mean": 0.0,
|
| 1566 |
+
"clip_ratio/low_mean": 0.0,
|
| 1567 |
+
"clip_ratio/low_min": 0.0,
|
| 1568 |
+
"clip_ratio/region_mean": 0.0,
|
| 1569 |
+
"epoch": 0.0005569007263922518,
|
| 1570 |
+
"grad_norm": 0.4252584937775968,
|
| 1571 |
+
"kl": 1.2265625,
|
| 1572 |
+
"learning_rate": 3e-06,
|
| 1573 |
+
"loss": 0.017,
|
| 1574 |
+
"step": 92
|
| 1575 |
+
},
|
| 1576 |
+
{
|
| 1577 |
+
"clip_ratio/high_max": 0.0,
|
| 1578 |
+
"clip_ratio/high_mean": 0.0,
|
| 1579 |
+
"clip_ratio/low_mean": 0.0,
|
| 1580 |
+
"clip_ratio/low_min": 0.0,
|
| 1581 |
+
"clip_ratio/region_mean": 0.0,
|
| 1582 |
+
"completions/clipped_ratio": 0.125,
|
| 1583 |
+
"completions/max_length": 2607.0,
|
| 1584 |
+
"completions/max_terminated_length": 2607.0,
|
| 1585 |
+
"completions/mean_length": 1312.6875,
|
| 1586 |
+
"completions/mean_terminated_length": 1500.2142857142858,
|
| 1587 |
+
"completions/min_length": 0.0,
|
| 1588 |
+
"completions/min_terminated_length": 442.0,
|
| 1589 |
+
"epoch": 0.000562953995157385,
|
| 1590 |
+
"grad_norm": 1.1205239401676392,
|
| 1591 |
+
"kl": 1.189453125,
|
| 1592 |
+
"learning_rate": 3e-06,
|
| 1593 |
+
"loss": -0.0069,
|
| 1594 |
+
"num_tokens": 739234.0,
|
| 1595 |
+
"reward": 0.033249590545892715,
|
| 1596 |
+
"reward_std": 0.13299836218357086,
|
| 1597 |
+
"rewards/avg_thinking_length_func": 151.1875,
|
| 1598 |
+
"rewards/correct_answer_reward_func": 0.0625,
|
| 1599 |
+
"rewards/efficient_thinking_reward_func": 0.4546527628842526,
|
| 1600 |
+
"rewards/format_reward_func": 0.9666666984558105,
|
| 1601 |
+
"rewards/num_xml_reward_func": 1.2395833730697632,
|
| 1602 |
+
"rewards/tool_execution_reward_func": 0.9375,
|
| 1603 |
+
"step": 93
|
| 1604 |
+
},
|
| 1605 |
+
{
|
| 1606 |
+
"clip_ratio/high_max": 0.0,
|
| 1607 |
+
"clip_ratio/high_mean": 0.0,
|
| 1608 |
+
"clip_ratio/low_mean": 0.0,
|
| 1609 |
+
"clip_ratio/low_min": 0.0,
|
| 1610 |
+
"clip_ratio/region_mean": 0.0,
|
| 1611 |
+
"epoch": 0.0005690072639225181,
|
| 1612 |
+
"grad_norm": 0.6129371822076389,
|
| 1613 |
+
"kl": 1.0341796875,
|
| 1614 |
+
"learning_rate": 3e-06,
|
| 1615 |
+
"loss": -0.0072,
|
| 1616 |
+
"step": 94
|
| 1617 |
+
},
|
| 1618 |
+
{
|
| 1619 |
+
"clip_ratio/high_max": 0.0,
|
| 1620 |
+
"clip_ratio/high_mean": 0.0,
|
| 1621 |
+
"clip_ratio/low_mean": 0.0,
|
| 1622 |
+
"clip_ratio/low_min": 0.0,
|
| 1623 |
+
"clip_ratio/region_mean": 0.0,
|
| 1624 |
+
"epoch": 0.0005750605326876513,
|
| 1625 |
+
"grad_norm": 0.30285345268245273,
|
| 1626 |
+
"kl": 0.8857421875,
|
| 1627 |
+
"learning_rate": 3e-06,
|
| 1628 |
+
"loss": -0.0076,
|
| 1629 |
+
"step": 95
|
| 1630 |
+
},
|
| 1631 |
+
{
|
| 1632 |
+
"clip_ratio/high_max": 0.0,
|
| 1633 |
+
"clip_ratio/high_mean": 0.0,
|
| 1634 |
+
"clip_ratio/low_mean": 0.0,
|
| 1635 |
+
"clip_ratio/low_min": 0.0,
|
| 1636 |
+
"clip_ratio/region_mean": 0.0,
|
| 1637 |
+
"epoch": 0.0005811138014527845,
|
| 1638 |
+
"grad_norm": 0.2502782554540184,
|
| 1639 |
+
"kl": 0.798828125,
|
| 1640 |
+
"learning_rate": 3e-06,
|
| 1641 |
+
"loss": -0.0079,
|
| 1642 |
+
"step": 96
|
| 1643 |
+
},
|
| 1644 |
+
{
|
| 1645 |
+
"clip_ratio/high_max": 0.0,
|
| 1646 |
+
"clip_ratio/high_mean": 0.0,
|
| 1647 |
+
"clip_ratio/low_mean": 0.0,
|
| 1648 |
+
"clip_ratio/low_min": 0.0,
|
| 1649 |
+
"clip_ratio/region_mean": 0.0,
|
| 1650 |
+
"completions/clipped_ratio": 0.0,
|
| 1651 |
+
"completions/max_length": 2939.0,
|
| 1652 |
+
"completions/max_terminated_length": 2939.0,
|
| 1653 |
+
"completions/mean_length": 1035.125,
|
| 1654 |
+
"completions/mean_terminated_length": 1035.125,
|
| 1655 |
+
"completions/min_length": 169.0,
|
| 1656 |
+
"completions/min_terminated_length": 169.0,
|
| 1657 |
+
"epoch": 0.0005871670702179177,
|
| 1658 |
+
"grad_norm": 5.13582049228313,
|
| 1659 |
+
"kl": 2.056640625,
|
| 1660 |
+
"learning_rate": 3e-06,
|
| 1661 |
+
"loss": 0.0147,
|
| 1662 |
+
"num_tokens": 763388.0,
|
| 1663 |
+
"reward": 0.18276317417621613,
|
| 1664 |
+
"reward_std": 0.3562992513179779,
|
| 1665 |
+
"rewards/avg_thinking_length_func": 61.21875,
|
| 1666 |
+
"rewards/correct_answer_reward_func": 0.5,
|
| 1667 |
+
"rewards/efficient_thinking_reward_func": 0.4562464008188011,
|
| 1668 |
+
"rewards/format_reward_func": 1.0,
|
| 1669 |
+
"rewards/num_xml_reward_func": 1.2916666269302368,
|
| 1670 |
+
"rewards/tool_execution_reward_func": 0.75,
|
| 1671 |
+
"step": 97
|
| 1672 |
+
},
|
| 1673 |
+
{
|
| 1674 |
+
"clip_ratio/high_max": 0.0,
|
| 1675 |
+
"clip_ratio/high_mean": 0.0,
|
| 1676 |
+
"clip_ratio/low_mean": 0.0,
|
| 1677 |
+
"clip_ratio/low_min": 0.0,
|
| 1678 |
+
"clip_ratio/region_mean": 0.0,
|
| 1679 |
+
"epoch": 0.0005932203389830508,
|
| 1680 |
+
"grad_norm": 0.8360603719298867,
|
| 1681 |
+
"kl": 1.5234375,
|
| 1682 |
+
"learning_rate": 3e-06,
|
| 1683 |
+
"loss": 0.0126,
|
| 1684 |
+
"step": 98
|
| 1685 |
+
},
|
| 1686 |
+
{
|
| 1687 |
+
"clip_ratio/high_max": 0.0,
|
| 1688 |
+
"clip_ratio/high_mean": 0.0,
|
| 1689 |
+
"clip_ratio/low_mean": 0.0,
|
| 1690 |
+
"clip_ratio/low_min": 0.0,
|
| 1691 |
+
"clip_ratio/region_mean": 0.0,
|
| 1692 |
+
"epoch": 0.000599273607748184,
|
| 1693 |
+
"grad_norm": 0.47398212103799814,
|
| 1694 |
+
"kl": 1.265625,
|
| 1695 |
+
"learning_rate": 3e-06,
|
| 1696 |
+
"loss": 0.0117,
|
| 1697 |
+
"step": 99
|
| 1698 |
+
},
|
| 1699 |
+
{
|
| 1700 |
+
"clip_ratio/high_max": 0.0,
|
| 1701 |
+
"clip_ratio/high_mean": 0.0,
|
| 1702 |
+
"clip_ratio/low_mean": 0.0,
|
| 1703 |
+
"clip_ratio/low_min": 0.0,
|
| 1704 |
+
"clip_ratio/region_mean": 0.0,
|
| 1705 |
+
"epoch": 0.0006053268765133172,
|
| 1706 |
+
"grad_norm": 0.4824867513848336,
|
| 1707 |
+
"kl": 1.189453125,
|
| 1708 |
+
"learning_rate": 3e-06,
|
| 1709 |
+
"loss": 0.0115,
|
| 1710 |
+
"step": 100
|
| 1711 |
+
}
|
| 1712 |
+
],
|
| 1713 |
+
"logging_steps": 1,
|
| 1714 |
+
"max_steps": 2000,
|
| 1715 |
+
"num_input_tokens_seen": 763388,
|
| 1716 |
+
"num_train_epochs": 1,
|
| 1717 |
+
"save_steps": 20,
|
| 1718 |
+
"stateful_callbacks": {
|
| 1719 |
+
"TrainerControl": {
|
| 1720 |
+
"args": {
|
| 1721 |
+
"should_epoch_stop": false,
|
| 1722 |
+
"should_evaluate": false,
|
| 1723 |
+
"should_log": false,
|
| 1724 |
+
"should_save": true,
|
| 1725 |
+
"should_training_stop": false
|
| 1726 |
+
},
|
| 1727 |
+
"attributes": {}
|
| 1728 |
+
}
|
| 1729 |
+
},
|
| 1730 |
+
"total_flos": 0.0,
|
| 1731 |
+
"train_batch_size": 1,
|
| 1732 |
+
"trial_name": null,
|
| 1733 |
+
"trial_params": null
|
| 1734 |
+
}
|
training_args.bin
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:ccc94e96d8759f5319cc1b34ffdc4a9111db8aa57df45d421dd02a88ba0cf1bd
|
| 3 |
+
size 8273
|
vocab.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|