Hugging Face's logo

whizzzzkid
/
mini122

Text Generation
Transformers
Safetensors
llama
conversational
text-generation-inference
Inference Endpoints
Model card Files Community
whizzzzkid commited on
Commit
887716b
•
1 Parent(s): cb9ae8b

Upload GemmaForCausalLM

Browse files
Files changed (4) hide show
  1. config.json +16 -17
  2. generation_config.json +5 -3
  3. model-00001-of-00002.safetensors +1 -1
  4. model-00002-of-00002.safetensors +1 -1
config.json CHANGED
@@ -1,29 +1,28 @@
1
  {
2
  "_name_or_path": "axalotl_ft",
3
  "architectures": [
4
- "StableLmForCausalLM"
5
  ],
 
6
  "attention_dropout": 0.0,
7
- "bos_token_id": 100289,
8
- "eos_token_id": 100290,
9
- "hidden_act": "silu",
10
- "hidden_dropout": 0.0,
11
  "hidden_size": 2048,
12
  "initializer_range": 0.02,
13
- "intermediate_size": 5632,
14
- "layer_norm_eps": 1e-05,
15
- "max_position_embeddings": 4096,
16
- "model_type": "stablelm",
17
- "num_attention_heads": 32,
18
- "num_hidden_layers": 24,
19
- "num_key_value_heads": 32,
20
- "partial_rotary_factor": 0.25,
21
  "rope_scaling": null,
22
- "rope_theta": 10000,
23
- "tie_word_embeddings": false,
24
  "torch_dtype": "bfloat16",
25
  "transformers_version": "4.38.2",
26
  "use_cache": false,
27
- "use_qkv_bias": true,
28
- "vocab_size": 100352
29
  }
 
1
  {
2
  "_name_or_path": "axalotl_ft",
3
  "architectures": [
4
+ "GemmaForCausalLM"
5
  ],
6
+ "attention_bias": false,
7
  "attention_dropout": 0.0,
8
+ "bos_token_id": 204,
9
+ "eos_token_id": 213,
10
+ "head_dim": 256,
11
+ "hidden_act": "gelu",
12
  "hidden_size": 2048,
13
  "initializer_range": 0.02,
14
+ "intermediate_size": 16384,
15
+ "max_position_embeddings": 8192,
16
+ "model_type": "gemma",
17
+ "num_attention_heads": 8,
18
+ "num_hidden_layers": 18,
19
+ "num_key_value_heads": 1,
20
+ "pad_token_id": 0,
21
+ "rms_norm_eps": 1e-06,
22
  "rope_scaling": null,
23
+ "rope_theta": 10000.0,
 
24
  "torch_dtype": "bfloat16",
25
  "transformers_version": "4.38.2",
26
  "use_cache": false,
27
+ "vocab_size": 256000
 
28
  }
generation_config.json CHANGED
@@ -1,6 +1,8 @@
1
  {
2
  "_from_model_config": true,
3
- "bos_token_id": 100257,
4
- "eos_token_id": 100257,
5
- "transformers_version": "4.38.2"
 
 
6
  }
 
1
  {
2
  "_from_model_config": true,
3
+ "bos_token_id": 204,
4
+ "eos_token_id": 213,
5
+ "pad_token_id": 0,
6
+ "transformers_version": "4.38.2",
7
+ "use_cache": false
8
  }
model-00001-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:34d1eb66fc1d745f9135727173641e5b01b739674850d700f2f0822c5d3ea3fe
3
  size 4945242264
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a6fd51fb0713a8c26dbf5812b69a4598aed51b8553ab714855f42601ef7b7cb6
3
  size 4945242264
model-00002-of-00002.safetensors CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:fa76d023fe4f9b5507e7b8c73b68cb91c74b23a4298193fc9ec54648868f00fa
3
  size 67121608
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bba9c83eccc5f711b413f09eae766175ccf8ad2822bbfde678f06ebe622afd50
3
  size 67121608