Trained with Unsloth
Browse files- config.json +38 -2
- pytorch_model-00001-of-00004.bin +1 -1
- pytorch_model-00002-of-00004.bin +2 -2
- pytorch_model-00003-of-00004.bin +1 -1
- pytorch_model-00004-of-00004.bin +1 -1
config.json
CHANGED
|
@@ -1,3 +1,39 @@
|
|
| 1 |
{
|
| 2 |
-
|
| 3 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
{
|
| 2 |
+
"_name_or_path": "unsloth/meta-llama-3.1-8b-unsloth-bnb-4bit",
|
| 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": 4096,
|
| 13 |
+
"initializer_range": 0.02,
|
| 14 |
+
"intermediate_size": 14336,
|
| 15 |
+
"max_position_embeddings": 131072,
|
| 16 |
+
"mlp_bias": false,
|
| 17 |
+
"model_type": "llama",
|
| 18 |
+
"num_attention_heads": 32,
|
| 19 |
+
"num_hidden_layers": 32,
|
| 20 |
+
"num_key_value_heads": 8,
|
| 21 |
+
"pad_token_id": 128004,
|
| 22 |
+
"pretraining_tp": 1,
|
| 23 |
+
"rms_norm_eps": 1e-05,
|
| 24 |
+
"rope_scaling": {
|
| 25 |
+
"factor": 8.0,
|
| 26 |
+
"high_freq_factor": 4.0,
|
| 27 |
+
"low_freq_factor": 1.0,
|
| 28 |
+
"original_max_position_embeddings": 8192,
|
| 29 |
+
"rope_type": "llama3"
|
| 30 |
+
},
|
| 31 |
+
"rope_theta": 500000.0,
|
| 32 |
+
"tie_word_embeddings": false,
|
| 33 |
+
"torch_dtype": "float16",
|
| 34 |
+
"transformers_version": "4.48.3",
|
| 35 |
+
"unsloth_fixed": true,
|
| 36 |
+
"unsloth_version": "2025.2.15",
|
| 37 |
+
"use_cache": true,
|
| 38 |
+
"vocab_size": 128256
|
| 39 |
+
}
|
pytorch_model-00001-of-00004.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4976718466
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:9627a3437372e4950e61f156479899fd41a5574377407d052f844d6bb51067bf
|
| 3 |
size 4976718466
|
pytorch_model-00002-of-00004.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
-
size
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:f355cbc78c22c90640681c977940a7361415a04e78b645847c755ad16540b908
|
| 3 |
+
size 4999827142
|
pytorch_model-00003-of-00004.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 4915939082
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:2ae260848bed399227808f26e7186587ec5e9b13c9a0e161a3c778d6497dfdb4
|
| 3 |
size 4915939082
|
pytorch_model-00004-of-00004.bin
CHANGED
|
@@ -1,3 +1,3 @@
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
-
oid sha256:
|
| 3 |
size 1168140873
|
|
|
|
| 1 |
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:452e7d020d51a567a4e7a846b5234699c534b07eb18a5d7684efc8617f31ce8f
|
| 3 |
size 1168140873
|