config files
Browse files- config.json +3 -36
- generation_config.json +3 -0
- model.safetensors.index.json +3 -0
- special_tokens_map.json +3 -0
- tokenizer.json +3 -0
- tokenizer_config.json +3 -0
- trainer_state.json +3 -0
config.json
CHANGED
@@ -1,36 +1,3 @@
|
|
1 |
-
|
2 |
-
|
3 |
-
|
4 |
-
"Gemma2ForCausalLM"
|
5 |
-
],
|
6 |
-
"attention_bias": false,
|
7 |
-
"attention_dropout": 0.0,
|
8 |
-
"attn_logit_softcapping": 50.0,
|
9 |
-
"bos_token_id": 2,
|
10 |
-
"cache_implementation": "hybrid",
|
11 |
-
"eos_token_id": [
|
12 |
-
1,
|
13 |
-
107
|
14 |
-
],
|
15 |
-
"final_logit_softcapping": 30.0,
|
16 |
-
"head_dim": 256,
|
17 |
-
"hidden_act": "gelu_pytorch_tanh",
|
18 |
-
"hidden_activation": "gelu_pytorch_tanh",
|
19 |
-
"hidden_size": 2304,
|
20 |
-
"initializer_range": 0.02,
|
21 |
-
"intermediate_size": 9216,
|
22 |
-
"max_position_embeddings": 8192,
|
23 |
-
"model_type": "gemma2",
|
24 |
-
"num_attention_heads": 8,
|
25 |
-
"num_hidden_layers": 26,
|
26 |
-
"num_key_value_heads": 4,
|
27 |
-
"pad_token_id": 0,
|
28 |
-
"query_pre_attn_scalar": 256,
|
29 |
-
"rms_norm_eps": 1e-06,
|
30 |
-
"rope_theta": 10000.0,
|
31 |
-
"sliding_window": 4096,
|
32 |
-
"torch_dtype": "bfloat16",
|
33 |
-
"transformers_version": "4.44.0",
|
34 |
-
"use_cache": false,
|
35 |
-
"vocab_size": 256000
|
36 |
-
}
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:e6b949d746d79773d4398e37a16bc22eff502c5420d5e03b68371bc3f21dff42
|
3 |
+
size 882
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
generation_config.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:93448ef46072177677ac4a9864d38e060d78376f0940409bb36f7cfb09bb1b3a
|
3 |
+
size 187
|
model.safetensors.index.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:ada0043f3e3b2e5ab2f445cad9c0fbbf9d91ad444675e6a82b822591c63abf5a
|
3 |
+
size 24223
|
special_tokens_map.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:baec30ea10906f16adb8c18af7a34023002c1746542612b8b41c9f09e1351351
|
3 |
+
size 636
|
tokenizer.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3f289bc05132635a8bc7aca7aa21255efd5e18f3710f43e3cdb96bcd41be4922
|
3 |
+
size 17525357
|
tokenizer_config.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:21e24ff1691d549c761afdd878eff196ac8d89df4e311b7ce3fa4676a1f4063f
|
3 |
+
size 46919
|
trainer_state.json
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:b8d6e05a242806385fdc8c6c61d49704c2ac5ac87a2ce615e97c67aff9b92108
|
3 |
+
size 1028067
|