nayemabs commited on
Commit
ba34281
·
verified ·
1 Parent(s): ee6948b

Update adapter_config.json

Browse files
Files changed (1) hide show
  1. adapter_config.json +5 -39
adapter_config.json CHANGED
@@ -1,38 +1,4 @@
1
  {
2
- "_name_or_path": "BanglaLLM/BanglaLLama-3.2-3b-unlop-culturax-base-v0.0.3",
3
- "architectures": [
4
- "LlamaForCausalLM"
5
- ],
6
- "attention_bias": false,
7
- "attention_dropout": 0.0,
8
- "bos_token_id": 128000,
9
- "eos_token_id": 128001,
10
- "head_dim": 128,
11
- "hidden_act": "silu",
12
- "hidden_size": 3072,
13
- "initializer_range": 0.02,
14
- "intermediate_size": 8192,
15
- "max_position_embeddings": 131072,
16
- "mlp_bias": false,
17
- "model_type": "llama",
18
- "num_attention_heads": 24,
19
- "num_hidden_layers": 28,
20
- "num_key_value_heads": 8,
21
- "pretraining_tp": 1,
22
- "rms_norm_eps": 1e-05,
23
- "rope_scaling": {
24
- "factor": 32.0,
25
- "high_freq_factor": 4.0,
26
- "low_freq_factor": 1.0,
27
- "original_max_position_embeddings": 8192,
28
- "rope_type": "llama3"
29
- },
30
- "rope_theta": 500000.0,
31
- "tie_word_embeddings": true,
32
- "torch_dtype": "bfloat16",
33
- "transformers_version": "4.45.2",
34
- "use_cache": true,
35
- "vocab_size": 128256,
36
  "alpha_pattern": {},
37
  "auto_mapping": null,
38
  "base_model_name_or_path": "BanglaLLM/BanglaLLama-3.2-3b-bangla-alpaca-orca-instruct-v0.0.1",
@@ -54,15 +20,15 @@
54
  "rank_pattern": {},
55
  "revision": null,
56
  "target_modules": [
57
- "q_proj",
58
  "v_proj",
59
  "o_proj",
60
- "down_proj",
61
- "up_proj",
62
  "gate_proj",
63
- "k_proj"
 
 
 
64
  ],
65
  "task_type": "CAUSAL_LM",
66
  "use_dora": false,
67
  "use_rslora": false
68
- }
 
1
  {
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
2
  "alpha_pattern": {},
3
  "auto_mapping": null,
4
  "base_model_name_or_path": "BanglaLLM/BanglaLLama-3.2-3b-bangla-alpaca-orca-instruct-v0.0.1",
 
20
  "rank_pattern": {},
21
  "revision": null,
22
  "target_modules": [
 
23
  "v_proj",
24
  "o_proj",
 
 
25
  "gate_proj",
26
+ "up_proj",
27
+ "k_proj",
28
+ "q_proj",
29
+ "down_proj"
30
  ],
31
  "task_type": "CAUSAL_LM",
32
  "use_dora": false,
33
  "use_rslora": false
34
+ }