ljq34952 commited on
Commit
10c1451
·
verified ·
1 Parent(s): 11b4bbb

Add files using upload-large-folder tool

Browse files
Files changed (47) hide show
  1. Qwen2.5-1.5B-R8-s0.0225-a0.0025/config.yaml +43 -0
  2. Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/custom/gsm8k-hard.json +3 -0
  3. Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/custom/multiarith.json +3 -0
  4. Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/custom/summary.json +3 -0
  5. Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/gsm8k.json +3 -0
  6. Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/svamp.json +3 -0
  7. Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/tensorboard/gsm8k/events.out.tfevents.1779967467.liu-SGASL90-002.2943492.0 +3 -0
  8. Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/tensorboard/svamp/events.out.tfevents.1779967497.liu-SGASL90-002.2944596.0 +3 -0
  9. Qwen2.5-1.5B-R8-s0.0225-a0.0025/history.json +3 -0
  10. Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/added_tokens.json +3 -0
  11. Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/chat_template.jinja +54 -0
  12. Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/config.json +3 -0
  13. Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/merges.txt +0 -0
  14. Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/model.safetensors +3 -0
  15. Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/special_tokens_map.json +3 -0
  16. Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/tokenizer.json +3 -0
  17. Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/tokenizer_config.json +3 -0
  18. Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/vocab.json +3 -0
  19. Qwen2.5-1.5B-R8-s0.0225-a0.0025/tensorboard/events.out.tfevents.1779809854.liu-SGASL90-002.4135416.0 +3 -0
  20. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/config.yaml +45 -0
  21. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/history.json +3 -0
  22. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/added_tokens.json +3 -0
  23. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/chat_template.jinja +54 -0
  24. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/config.json +3 -0
  25. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/merges.txt +0 -0
  26. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/model.safetensors +3 -0
  27. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/special_tokens_map.json +3 -0
  28. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/tokenizer.json +3 -0
  29. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/tokenizer_config.json +3 -0
  30. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/vocab.json +3 -0
  31. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/tensorboard/events.out.tfevents.1779969335.liu-SGASL90-002.2964382.0 +3 -0
  32. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/tensorboard/events.out.tfevents.1780038299.liu-SGASL90-002.2325047.0 +3 -0
  33. Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/tensorboard/events.out.tfevents.1780043470.liu-SGASL90-002.2722148.0 +3 -0
  34. Qwen2.5-1.5B-R8-s0.0225-a0.0035/config.yaml +45 -0
  35. Qwen2.5-1.5B-R8-s0.0225-a0.0035/history.json +3 -0
  36. Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/added_tokens.json +3 -0
  37. Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/chat_template.jinja +54 -0
  38. Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/config.json +3 -0
  39. Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/merges.txt +0 -0
  40. Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/model.safetensors +3 -0
  41. Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/special_tokens_map.json +3 -0
  42. Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/tokenizer.json +3 -0
  43. Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/tokenizer_config.json +3 -0
  44. Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/vocab.json +3 -0
  45. Qwen2.5-1.5B-R8-s0.0225-a0.0035/tensorboard/events.out.tfevents.1779969334.liu-SGASL90-002.2964317.0 +3 -0
  46. Qwen2.5-1.5B-R8-s0.0225-a0.0035/tensorboard/events.out.tfevents.1780038295.liu-SGASL90-002.2325043.0 +3 -0
  47. Qwen2.5-1.5B-R8-s0.0225-a0.0035/tensorboard/events.out.tfevents.1780043163.liu-SGASL90-002.2691513.0 +3 -0
Qwen2.5-1.5B-R8-s0.0225-a0.0025/config.yaml ADDED
@@ -0,0 +1,43 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ es:
2
+ n_epochs: 10
3
+ batch_size: 100
4
+ sigma: 0.0225
5
+ alpha: 0.0025
6
+ pop_size: 60
7
+ reward_normalization: zscore
8
+ schedule_mode: keep_tail_per_epoch
9
+ reward_normalization_ddof: 0
10
+ reward_normalization_eps: 1.0e-08
11
+ raw_train_size: 7473
12
+ effective_train_size: 7473
13
+ max_steps: 750
14
+ config_version: 2
15
+ method: es
16
+ finetune:
17
+ type: lora
18
+ lora_rank: 8
19
+ lora_alpha: 8
20
+ target_modules:
21
+ - q_proj
22
+ - k_proj
23
+ - v_proj
24
+ - o_proj
25
+ - gate_proj
26
+ - up_proj
27
+ - down_proj
28
+ method_params: {}
29
+ output:
30
+ seed: 42
31
+ dir: /home/bayp/ljq/ES4LLM/curves/lora/r8/20260526_233608_4135366/sigma0.0225_alpha0.0025
32
+ dtype: bfloat16
33
+ benchmark: gsm8k
34
+ gsm8k_path: /home/bayp/EsSftSVD/data/dataset/gsm8k
35
+ checkpoint:
36
+ save_freq: 50
37
+ export_merged_freq: 50
38
+ generation:
39
+ max_new_tokens: 2048
40
+ format_reward: 0.1
41
+ model: /home/bayp/EsSftSVD/data/model/Qwen2.5-1.5B-Instruct
42
+ runtime:
43
+ model: /home/bayp/EsSftSVD/data/model/Qwen2.5-1.5B-Instruct
Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/custom/gsm8k-hard.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb7fd4a02fc49fdd9d793105893d6dc682b9be3b328b3ae3df177b3d42f2139c
3
+ size 2222095
Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/custom/multiarith.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:df86227c4233e1e5c1f58623770ed3b69de10fda104bda70fbe0ab9a85f70cef
3
+ size 189500
Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/custom/summary.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6151d3e427b7a9b0893e7e27d374437dbdf82ea5718b1b3cd458a1daf8b90549
3
+ size 189423
Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/gsm8k.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:27e26c0770878cbb8bcce969612e46b5bedb1aaf2eac2d9d244d5bea3bb9649a
3
+ size 2009826
Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/svamp.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9f8656a2ad0deeac235c8877e71c74e5f0606fa746e7c81dca113938bc9c0c34
3
+ size 366520
Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/tensorboard/gsm8k/events.out.tfevents.1779967467.liu-SGASL90-002.2943492.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:239ea820188de36f259d41fbb35e4657094a94b0355a9388083ad72a35268a23
3
+ size 791
Qwen2.5-1.5B-R8-s0.0225-a0.0025/eval_r8_s0.0225_a0.0025/tensorboard/svamp/events.out.tfevents.1779967497.liu-SGASL90-002.2944596.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:942566d1aa3d6cc6686c61a1656806e31781841963e822f4786a5c554f603af5
3
+ size 791
Qwen2.5-1.5B-R8-s0.0225-a0.0025/history.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3f2d8af837c0ad3666b08e7ab95c3ba9aaa3a97d5e4a26b8b09014ce9ffe5887
3
+ size 100238
Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58b54bbe36fc752f79a24a271ef66a0a0830054b4dfad94bde757d851968060b
3
+ size 605
Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/chat_template.jinja ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
7
+ {%- endif %}
8
+ {{- "\n\n# 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>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\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" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- endif %}
Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/config.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c056f4a27d910748710a6cfb66061c3a2d36ca5b2ff0c7df9380cd17bf9e2d2b
3
+ size 1400
Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c29ede7766647aab1b49fe7464cdb7e1f39c3920f1a3e6253f280f03a6759300
3
+ size 3554214720
Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/special_tokens_map.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76862e765266b85aa9459767e33cbaf13970f327a0e88d1c65846c2ddd3a1ecd
3
+ size 613
Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
3
+ size 11421896
Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/tokenizer_config.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a04a9d7d4a62b28482bdfe726c122756de85714fb64166ace92ae75b8f57614
3
+ size 4686
Qwen2.5-1.5B-R8-s0.0225-a0.0025/merged/vocab.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca10d7e9fb3ed18575dd1e277a2579c16d108e32f27439684afa0e10b1440910
3
+ size 2776833
Qwen2.5-1.5B-R8-s0.0225-a0.0025/tensorboard/events.out.tfevents.1779809854.liu-SGASL90-002.4135416.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7a76c78b1f9b19b318f604424efab4dbe5efc8b3ec623e57ae6d95d91c8e4baf
3
+ size 127009
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/config.yaml ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ es:
2
+ n_epochs: 10
3
+ batch_size: 100
4
+ sigma: 0.0225
5
+ alpha: 0.0025
6
+ pop_size: 60
7
+ reward_normalization: zscore
8
+ schedule_mode: keep_tail_per_epoch
9
+ reward_normalization_ddof: 0
10
+ reward_normalization_eps: 1.0e-08
11
+ raw_train_size: 7473
12
+ effective_train_size: 7473
13
+ max_steps: 750
14
+ config_version: 2
15
+ method: es
16
+ finetune:
17
+ type: lora
18
+ lora_rank: 8
19
+ lora_alpha: 8
20
+ target_modules:
21
+ - q_proj
22
+ - k_proj
23
+ - v_proj
24
+ - o_proj
25
+ - gate_proj
26
+ - up_proj
27
+ - down_proj
28
+ method_params: {}
29
+ output:
30
+ seed: 42
31
+ dir: /home/bayp/ljq/ES4LLM/curves/lora/r8/20260528_195410_2964267/sigma0.0225_alpha0.0025
32
+ dtype: bfloat16
33
+ benchmark: gsm8k
34
+ gsm8k_path: /home/bayp/EsSftSVD/data/dataset/gsm8k
35
+ checkpoint:
36
+ save_freq: 50
37
+ export_merged_freq: 50
38
+ resume_from: /home/bayp/ljq/ES4LLM/curves/lora/r8/20260528_195410_2964267/sigma0.0225_alpha0.0025/Qwen2.5-1.5B-Instruct/sigma0.0225_alpha0.0025_N60_ep10_s42_20260528_195416/checkpoints/latest.pt
39
+ generation:
40
+ max_new_tokens: 2048
41
+ format_reward: 0.1
42
+ merged_model: /data/ES4LLM/archive/20260528_195410_2964267_active_merged_checkpoints/sigma0.0225_alpha0.0025__Qwen2.5-1.5B-Instruct__sigma0.0225_alpha0.0025_N60_ep10_s42_20260528_195416__merged_checkpoints__step_400
43
+ model: /home/bayp/EsSftSVD/data/model/Qwen2.5-1.5B-Instruct
44
+ runtime:
45
+ model: /data/ES4LLM/archive/20260528_195410_2964267_active_merged_checkpoints/sigma0.0225_alpha0.0025__Qwen2.5-1.5B-Instruct__sigma0.0225_alpha0.0025_N60_ep10_s42_20260528_195416__merged_checkpoints__step_400
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/history.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f2efc14fc01b8e87526e4f3ff86c3d525caa7ed30f75fd8dd93a140b2f74cbf1
3
+ size 100242
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58b54bbe36fc752f79a24a271ef66a0a0830054b4dfad94bde757d851968060b
3
+ size 605
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/chat_template.jinja ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
7
+ {%- endif %}
8
+ {{- "\n\n# 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>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\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" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- endif %}
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/config.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c056f4a27d910748710a6cfb66061c3a2d36ca5b2ff0c7df9380cd17bf9e2d2b
3
+ size 1400
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a40d35877a78d7563533ca917e750a2b860d00bc9f4d0d66e37191e5f8c6b77
3
+ size 3554214720
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/special_tokens_map.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76862e765266b85aa9459767e33cbaf13970f327a0e88d1c65846c2ddd3a1ecd
3
+ size 613
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
3
+ size 11421896
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/tokenizer_config.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a04a9d7d4a62b28482bdfe726c122756de85714fb64166ace92ae75b8f57614
3
+ size 4686
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/merged/vocab.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca10d7e9fb3ed18575dd1e277a2579c16d108e32f27439684afa0e10b1440910
3
+ size 2776833
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/tensorboard/events.out.tfevents.1779969335.liu-SGASL90-002.2964382.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ea4d00365cf8b0b9d58d881699aee8be9fae0fc106b50bd5a30f75cb341fa35
3
+ size 69859
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/tensorboard/events.out.tfevents.1780038299.liu-SGASL90-002.2325047.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f3429670bb6581d6e35d0a2e948b75965eca30bd44004cda67a5a98f47d253a9
3
+ size 3413
Qwen2.5-1.5B-R8-s0.0225-a0.0025_1/tensorboard/events.out.tfevents.1780043470.liu-SGASL90-002.2722148.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9727bd93d384e87f8ffc59c494ccc7f67b3d98c2cc14c0138e3102085389ed13
3
+ size 60736
Qwen2.5-1.5B-R8-s0.0225-a0.0035/config.yaml ADDED
@@ -0,0 +1,45 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ es:
2
+ n_epochs: 10
3
+ batch_size: 100
4
+ sigma: 0.0225
5
+ alpha: 0.0035
6
+ pop_size: 60
7
+ reward_normalization: zscore
8
+ schedule_mode: keep_tail_per_epoch
9
+ reward_normalization_ddof: 0
10
+ reward_normalization_eps: 1.0e-08
11
+ raw_train_size: 7473
12
+ effective_train_size: 7473
13
+ max_steps: 750
14
+ config_version: 2
15
+ method: es
16
+ finetune:
17
+ type: lora
18
+ lora_rank: 8
19
+ lora_alpha: 8
20
+ target_modules:
21
+ - q_proj
22
+ - k_proj
23
+ - v_proj
24
+ - o_proj
25
+ - gate_proj
26
+ - up_proj
27
+ - down_proj
28
+ method_params: {}
29
+ output:
30
+ seed: 42
31
+ dir: /home/bayp/ljq/ES4LLM/curves/lora/r8/20260528_195410_2964267/sigma0.0225_alpha0.0035
32
+ dtype: bfloat16
33
+ benchmark: gsm8k
34
+ gsm8k_path: /home/bayp/EsSftSVD/data/dataset/gsm8k
35
+ checkpoint:
36
+ save_freq: 50
37
+ export_merged_freq: 50
38
+ resume_from: /home/bayp/ljq/ES4LLM/curves/lora/r8/20260528_195410_2964267/sigma0.0225_alpha0.0035/Qwen2.5-1.5B-Instruct/sigma0.0225_alpha0.0035_N60_ep10_s42_20260528_195416/checkpoints/latest.pt
39
+ generation:
40
+ max_new_tokens: 2048
41
+ format_reward: 0.1
42
+ merged_model: /data/ES4LLM/archive/20260528_195410_2964267_active_merged_checkpoints/sigma0.0225_alpha0.0035__Qwen2.5-1.5B-Instruct__sigma0.0225_alpha0.0035_N60_ep10_s42_20260528_195416__merged_checkpoints__step_400
43
+ model: /home/bayp/EsSftSVD/data/model/Qwen2.5-1.5B-Instruct
44
+ runtime:
45
+ model: /data/ES4LLM/archive/20260528_195410_2964267_active_merged_checkpoints/sigma0.0225_alpha0.0035__Qwen2.5-1.5B-Instruct__sigma0.0225_alpha0.0035_N60_ep10_s42_20260528_195416__merged_checkpoints__step_400
Qwen2.5-1.5B-R8-s0.0225-a0.0035/history.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:683ab0a531a50c3ac44ed0dcf4aca61c2c8c9297b65082efffcc5d2d3554ecb0
3
+ size 100192
Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/added_tokens.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:58b54bbe36fc752f79a24a271ef66a0a0830054b4dfad94bde757d851968060b
3
+ size 605
Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/chat_template.jinja ADDED
@@ -0,0 +1,54 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {%- if tools %}
2
+ {{- '<|im_start|>system\n' }}
3
+ {%- if messages[0]['role'] == 'system' %}
4
+ {{- messages[0]['content'] }}
5
+ {%- else %}
6
+ {{- 'You are Qwen, created by Alibaba Cloud. You are a helpful assistant.' }}
7
+ {%- endif %}
8
+ {{- "\n\n# 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>" }}
9
+ {%- for tool in tools %}
10
+ {{- "\n" }}
11
+ {{- tool | tojson }}
12
+ {%- endfor %}
13
+ {{- "\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" }}
14
+ {%- else %}
15
+ {%- if messages[0]['role'] == 'system' %}
16
+ {{- '<|im_start|>system\n' + messages[0]['content'] + '<|im_end|>\n' }}
17
+ {%- else %}
18
+ {{- '<|im_start|>system\nYou are Qwen, created by Alibaba Cloud. You are a helpful assistant.<|im_end|>\n' }}
19
+ {%- endif %}
20
+ {%- endif %}
21
+ {%- for message in messages %}
22
+ {%- if (message.role == "user") or (message.role == "system" and not loop.first) or (message.role == "assistant" and not message.tool_calls) %}
23
+ {{- '<|im_start|>' + message.role + '\n' + message.content + '<|im_end|>' + '\n' }}
24
+ {%- elif message.role == "assistant" %}
25
+ {{- '<|im_start|>' + message.role }}
26
+ {%- if message.content %}
27
+ {{- '\n' + message.content }}
28
+ {%- endif %}
29
+ {%- for tool_call in message.tool_calls %}
30
+ {%- if tool_call.function is defined %}
31
+ {%- set tool_call = tool_call.function %}
32
+ {%- endif %}
33
+ {{- '\n<tool_call>\n{"name": "' }}
34
+ {{- tool_call.name }}
35
+ {{- '", "arguments": ' }}
36
+ {{- tool_call.arguments | tojson }}
37
+ {{- '}\n</tool_call>' }}
38
+ {%- endfor %}
39
+ {{- '<|im_end|>\n' }}
40
+ {%- elif message.role == "tool" %}
41
+ {%- if (loop.index0 == 0) or (messages[loop.index0 - 1].role != "tool") %}
42
+ {{- '<|im_start|>user' }}
43
+ {%- endif %}
44
+ {{- '\n<tool_response>\n' }}
45
+ {{- message.content }}
46
+ {{- '\n</tool_response>' }}
47
+ {%- if loop.last or (messages[loop.index0 + 1].role != "tool") %}
48
+ {{- '<|im_end|>\n' }}
49
+ {%- endif %}
50
+ {%- endif %}
51
+ {%- endfor %}
52
+ {%- if add_generation_prompt %}
53
+ {{- '<|im_start|>assistant\n' }}
54
+ {%- endif %}
Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/config.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c056f4a27d910748710a6cfb66061c3a2d36ca5b2ff0c7df9380cd17bf9e2d2b
3
+ size 1400
Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/merges.txt ADDED
The diff for this file is too large to render. See raw diff
 
Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c21ddbdb98b98b30e62e726497323c51501cf2dd11bde55697cd173034c71d0
3
+ size 3554214720
Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/special_tokens_map.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:76862e765266b85aa9459767e33cbaf13970f327a0e88d1c65846c2ddd3a1ecd
3
+ size 613
Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/tokenizer.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9c5ae00e602b8860cbd784ba82a8aa14e8feecec692e7076590d014d7b7fdafa
3
+ size 11421896
Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/tokenizer_config.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0a04a9d7d4a62b28482bdfe726c122756de85714fb64166ace92ae75b8f57614
3
+ size 4686
Qwen2.5-1.5B-R8-s0.0225-a0.0035/merged/vocab.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ca10d7e9fb3ed18575dd1e277a2579c16d108e32f27439684afa0e10b1440910
3
+ size 2776833
Qwen2.5-1.5B-R8-s0.0225-a0.0035/tensorboard/events.out.tfevents.1779969334.liu-SGASL90-002.2964317.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:64da78f99b50d1c3e8bc1ceaa3ee0b7e10d58b5153096104fc10276e80e0a54e
3
+ size 70024
Qwen2.5-1.5B-R8-s0.0225-a0.0035/tensorboard/events.out.tfevents.1780038295.liu-SGASL90-002.2325043.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2b108433ba693e6f41554854636b175102895a9d52491a00b1da1ff65aae6f95
3
+ size 3413
Qwen2.5-1.5B-R8-s0.0225-a0.0035/tensorboard/events.out.tfevents.1780043163.liu-SGASL90-002.2691513.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:715d53e875f67a325df3c43858a4787d1e2b3ef0320d1920532245228163dc6e
3
+ size 60736