Uploaded config.json for inference api to work
Browse files- config.json +47 -0
config.json
ADDED
|
@@ -0,0 +1,47 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_name_or_path": "/workspace/process/huggingfaceh4_zephyr-7b-beta/source",
|
| 3 |
+
"architectures": [
|
| 4 |
+
"MistralForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"bos_token_id": 1,
|
| 7 |
+
"eos_token_id": 2,
|
| 8 |
+
"hidden_act": "silu",
|
| 9 |
+
"hidden_size": 4096,
|
| 10 |
+
"initializer_range": 0.02,
|
| 11 |
+
"intermediate_size": 14336,
|
| 12 |
+
"max_position_embeddings": 32768,
|
| 13 |
+
"model_type": "mistral",
|
| 14 |
+
"num_attention_heads": 32,
|
| 15 |
+
"num_hidden_layers": 32,
|
| 16 |
+
"num_key_value_heads": 8,
|
| 17 |
+
"pad_token_id": 2,
|
| 18 |
+
"pretraining_tp": 1,
|
| 19 |
+
"quantization_config": {
|
| 20 |
+
"batch_size": 1,
|
| 21 |
+
"bits": 4,
|
| 22 |
+
"block_name_to_quantize": "model.layers",
|
| 23 |
+
"damp_percent": 0.1,
|
| 24 |
+
"desc_act": true,
|
| 25 |
+
"disable_exllama": false,
|
| 26 |
+
"group_size": 128,
|
| 27 |
+
"max_input_length": null,
|
| 28 |
+
"model_seqlen": 4096,
|
| 29 |
+
"module_name_preceding_first_block": [
|
| 30 |
+
"model.embed_tokens"
|
| 31 |
+
],
|
| 32 |
+
"pad_token_id": null,
|
| 33 |
+
"quant_method": "gptq",
|
| 34 |
+
"sym": true,
|
| 35 |
+
"tokenizer": null,
|
| 36 |
+
"true_sequential": true,
|
| 37 |
+
"use_cuda_fp16": true
|
| 38 |
+
},
|
| 39 |
+
"rms_norm_eps": 1e-05,
|
| 40 |
+
"rope_theta": 10000.0,
|
| 41 |
+
"sliding_window": 4096,
|
| 42 |
+
"tie_word_embeddings": false,
|
| 43 |
+
"torch_dtype": "float16",
|
| 44 |
+
"transformers_version": "4.34.1",
|
| 45 |
+
"use_cache": true,
|
| 46 |
+
"vocab_size": 32000
|
| 47 |
+
}
|