Commit
•
063f3c0
1
Parent(s):
7979aad
GPTQ commit model & tokenizer
Browse files- config.json +42 -0
- gptq_model-4bit-128g-00001-of-00004.safetensors +3 -0
- gptq_model-4bit-128g-00002-of-00004.safetensors +3 -0
- gptq_model-4bit-128g-00003-of-00004.safetensors +3 -0
- gptq_model-4bit-128g-00004-of-00004.safetensors +3 -0
- gptq_model-4bit-128g.safetensors.index.json +0 -0
- quantize_config.json +13 -0
config.json
ADDED
|
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "abacusai/Smaug-Llama-3-70B-Instruct",
|
| 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 |
+
"hidden_act": "silu",
|
| 11 |
+
"hidden_size": 8192,
|
| 12 |
+
"initializer_range": 0.02,
|
| 13 |
+
"intermediate_size": 28672,
|
| 14 |
+
"max_position_embeddings": 8192,
|
| 15 |
+
"model_type": "llama",
|
| 16 |
+
"num_attention_heads": 64,
|
| 17 |
+
"num_hidden_layers": 80,
|
| 18 |
+
"num_key_value_heads": 8,
|
| 19 |
+
"pretraining_tp": 1,
|
| 20 |
+
"quantization_config": {
|
| 21 |
+
"bits": 4,
|
| 22 |
+
"checkpoint_format": "gptq",
|
| 23 |
+
"damp_percent": 0.01,
|
| 24 |
+
"desc_act": true,
|
| 25 |
+
"group_size": 128,
|
| 26 |
+
"model_file_base_name": null,
|
| 27 |
+
"model_name_or_path": null,
|
| 28 |
+
"quant_method": "gptq",
|
| 29 |
+
"static_groups": false,
|
| 30 |
+
"sym": true,
|
| 31 |
+
"true_sequential": true
|
| 32 |
+
},
|
| 33 |
+
"rms_norm_eps": 1e-05,
|
| 34 |
+
"rope_scaling": null,
|
| 35 |
+
"rope_theta": 500000.0,
|
| 36 |
+
"tie_word_embeddings": false,
|
| 37 |
+
"torch_dtype": "float16",
|
| 38 |
+
"transformers_version": "4.40.1",
|
| 39 |
+
"use_cache": true,
|
| 40 |
+
"use_flash_attention_2": true,
|
| 41 |
+
"vocab_size": 128256
|
| 42 |
+
}
|
gptq_model-4bit-128g-00001-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0adfb2819768001755c7b771bc76aefae12738e1d0a677eadc8ee65866de13ae
|
| 3 |
+
size 9986976888
|
gptq_model-4bit-128g-00002-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3c0154cb745700109a458e968e2a2384820803792774fb0c9f68f4be0e14798c
|
| 3 |
+
size 9988075240
|
gptq_model-4bit-128g-00003-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:0d348a1de9ff8f92964edd6182e67bfe006af067120399b075a7a8610d4bd4f3
|
| 3 |
+
size 9909380848
|
gptq_model-4bit-128g-00004-of-00004.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:faaef9b8a53ad4f53997361c01b4f1dbcdb4100237f28c08dc3bd66214418e5d
|
| 3 |
+
size 9908529096
|
gptq_model-4bit-128g.safetensors.index.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
quantize_config.json
ADDED
|
@@ -0,0 +1,13 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bits": 4,
|
| 3 |
+
"group_size": 128,
|
| 4 |
+
"damp_percent": 0.01,
|
| 5 |
+
"desc_act": true,
|
| 6 |
+
"static_groups": false,
|
| 7 |
+
"sym": true,
|
| 8 |
+
"true_sequential": true,
|
| 9 |
+
"model_name_or_path": null,
|
| 10 |
+
"model_file_base_name": null,
|
| 11 |
+
"quant_method": "gptq",
|
| 12 |
+
"checkpoint_format": "gptq"
|
| 13 |
+
}
|