Hugging Face's logo

SeanLee97
/
bellm-llama-7b-nli

PEFT
TensorBoard
Safetensors
Model card Files Metrics Community
SeanLee97 commited on
Commit
c87410c
•
1 Parent(s): 6587a30

Training in progress, step 2800

Browse files
Files changed (6) hide show
  1. adapter_config.json +5 -5
  2. adapter_model.safetensors +2 -2
  3. angle.config +5 -5
  4. runs/Mar26_14-12-55_instance-20240226-074643/events.out.tfevents.1711462382.instance-20240226-074643.1699661.0 +2 -2
  5. runs/Mar27_06-18-35_instance-20240226-074643/events.out.tfevents.1711520320.instance-20240226-074643.2135606.0 +3 -0
  6. training_args.bin +1 -1
adapter_config.json CHANGED
@@ -15,17 +15,17 @@
15
  "megatron_core": "megatron.core",
16
  "modules_to_save": null,
17
  "peft_type": "LORA",
18
- "r": 128,
19
  "rank_pattern": {},
20
  "revision": null,
21
  "target_modules": [
22
- "up_proj",
23
- "q_proj",
24
  "v_proj",
25
  "k_proj",
26
- "gate_proj",
 
 
27
  "down_proj",
28
- "o_proj"
29
  ],
30
  "task_type": "CAUSAL_LM",
31
  "use_rslora": false
 
15
  "megatron_core": "megatron.core",
16
  "modules_to_save": null,
17
  "peft_type": "LORA",
18
+ "r": 64,
19
  "rank_pattern": {},
20
  "revision": null,
21
  "target_modules": [
 
 
22
  "v_proj",
23
  "k_proj",
24
+ "q_proj",
25
+ "up_proj",
26
+ "o_proj",
27
  "down_proj",
28
+ "gate_proj"
29
  ],
30
  "task_type": "CAUSAL_LM",
31
  "use_rslora": false
adapter_model.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fc0f2b70c08550a8f477209c70289eab2014f47d8c3dbb3b360e1b3df7962ba9
3
- size 1279323952
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fd698023e75d01002cb7d034e908ecc61ef3cb5fff66504923417c2f208f5ca4
3
+ size 639691872
angle.config CHANGED
@@ -5,18 +5,18 @@
5
  "pooling_strategy": "cls",
6
  "lora_config_kwargs": {
7
  "task_type": "CAUSAL_LM",
8
- "r": 128,
9
  "lora_alpha": 8,
10
  "lora_dropout": 0.05,
11
  "bias": "none",
12
  "target_modules": [
13
- "k_proj",
14
  "q_proj",
15
  "v_proj",
16
- "o_proj",
17
- "gate_proj",
18
  "down_proj",
19
- "up_proj"
 
 
20
  ]
21
  },
22
  "apply_lora": 1
 
5
  "pooling_strategy": "cls",
6
  "lora_config_kwargs": {
7
  "task_type": "CAUSAL_LM",
8
+ "r": 64,
9
  "lora_alpha": 8,
10
  "lora_dropout": 0.05,
11
  "bias": "none",
12
  "target_modules": [
13
+ "gate_proj",
14
  "q_proj",
15
  "v_proj",
 
 
16
  "down_proj",
17
+ "o_proj",
18
+ "up_proj",
19
+ "k_proj"
20
  ]
21
  },
22
  "apply_lora": 1
runs/Mar26_14-12-55_instance-20240226-074643/events.out.tfevents.1711462382.instance-20240226-074643.1699661.0 CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:06f59e724c39372178c20cc3da0361f799fbbb72262ba16066fd38e0816476bf
3
- size 9844
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c1a577d865830312bcb9f2f784beb91f902e0fbe2120317c892983e2ccd064c5
3
+ size 15262
runs/Mar27_06-18-35_instance-20240226-074643/events.out.tfevents.1711520320.instance-20240226-074643.2135606.0 ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4fb3352619fd83f6a9f75c4fa5e44796680c691a6b3a80d6908d1528848cfdb7
3
+ size 11110
training_args.bin CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:b8c41d6589f3f961edd3468c3af13c3c5dd8f49872630e845f622fd35b2314b2
3
  size 5048
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:877ab143cea4335808e6b29f3f8885626cedabdd74355a641c9500e407e9f5b6
3
  size 5048