BART-ender commited on
Commit
0cd638b
·
verified ·
1 Parent(s): fb8489c

Upload folder using huggingface_hub

Browse files
blue_cycle_1/README.md CHANGED
@@ -1,9 +1,9 @@
1
  ---
2
- base_model: unsloth/Qwen2.5-3B-Instruct-bnb-4bit
3
  library_name: peft
4
  pipeline_tag: text-generation
5
  tags:
6
- - base_model:adapter:unsloth/Qwen2.5-3B-Instruct-bnb-4bit
7
  - grpo
8
  - lora
9
  - transformers
 
1
  ---
2
+ base_model: unsloth/Qwen2.5-0.5B-Instruct-bnb-4bit
3
  library_name: peft
4
  pipeline_tag: text-generation
5
  tags:
6
+ - base_model:adapter:unsloth/Qwen2.5-0.5B-Instruct-bnb-4bit
7
  - grpo
8
  - lora
9
  - transformers
blue_cycle_1/adapter_config.json CHANGED
@@ -7,7 +7,7 @@
7
  "parent_library": "transformers.models.qwen2.modeling_qwen2",
8
  "unsloth_fixed": true
9
  },
10
- "base_model_name_or_path": "unsloth/Qwen2.5-3B-Instruct-bnb-4bit",
11
  "bias": "none",
12
  "corda_config": null,
13
  "ensure_weight_tying": false,
@@ -20,7 +20,7 @@
20
  "layers_pattern": null,
21
  "layers_to_transform": null,
22
  "loftq_config": {},
23
- "lora_alpha": 32,
24
  "lora_bias": false,
25
  "lora_dropout": 0,
26
  "lora_ga_config": null,
@@ -30,16 +30,11 @@
30
  "peft_type": "LORA",
31
  "peft_version": "0.19.1",
32
  "qalora_group_size": 16,
33
- "r": 16,
34
  "rank_pattern": {},
35
  "revision": null,
36
  "target_modules": [
37
  "q_proj",
38
- "down_proj",
39
- "gate_proj",
40
- "o_proj",
41
- "up_proj",
42
- "k_proj",
43
  "v_proj"
44
  ],
45
  "target_parameters": null,
 
7
  "parent_library": "transformers.models.qwen2.modeling_qwen2",
8
  "unsloth_fixed": true
9
  },
10
+ "base_model_name_or_path": "unsloth/Qwen2.5-0.5B-Instruct-bnb-4bit",
11
  "bias": "none",
12
  "corda_config": null,
13
  "ensure_weight_tying": false,
 
20
  "layers_pattern": null,
21
  "layers_to_transform": null,
22
  "loftq_config": {},
23
+ "lora_alpha": 16,
24
  "lora_bias": false,
25
  "lora_dropout": 0,
26
  "lora_ga_config": null,
 
30
  "peft_type": "LORA",
31
  "peft_version": "0.19.1",
32
  "qalora_group_size": 16,
33
+ "r": 8,
34
  "rank_pattern": {},
35
  "revision": null,
36
  "target_modules": [
37
  "q_proj",
 
 
 
 
 
38
  "v_proj"
39
  ],
40
  "target_parameters": null,
blue_cycle_1/adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:8851d689bbeba6aced1434b3a33338c1758f3907c145a95d87d9abe1929026a3
3
- size 119801528
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d81a8869746a3cce0238b90ae22f1555ede481ec8b1c4a6c5a4168189c996f2a
3
+ size 2175168
blue_cycle_1/best_metrics.json ADDED
@@ -0,0 +1,6 @@
 
 
 
 
 
 
 
1
+ {
2
+ "metric": "train/loss",
3
+ "best_loss": -0.01885230392217636,
4
+ "best_step": 155,
5
+ "agent_type": "blue"
6
+ }
blue_cycle_1/tokenizer_config.json CHANGED
@@ -8,7 +8,7 @@
8
  "is_local": false,
9
  "model_max_length": 32768,
10
  "pad_token": "<|PAD_TOKEN|>",
11
- "padding_side": "left",
12
  "split_special_tokens": false,
13
  "tokenizer_class": "Qwen2Tokenizer",
14
  "unk_token": null,
 
8
  "is_local": false,
9
  "model_max_length": 32768,
10
  "pad_token": "<|PAD_TOKEN|>",
11
+ "padding_side": "right",
12
  "split_special_tokens": false,
13
  "tokenizer_class": "Qwen2Tokenizer",
14
  "unk_token": null,