Upload 3 files
Browse files- config.json +56 -0
- mm_projector.bin +3 -0
- trainer_state.json +0 -0
config.json
ADDED
@@ -0,0 +1,56 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_name_or_path": "/leonardo_scratch/large/userexternal/fcocchi0/rag_mlmm/hf_models/llama_3_1/models--meta-llama--Meta-Llama-3.1-8B-Instruct/snapshots/07eb05b21d191a58c577b4a45982fe0c049d0693",
|
3 |
+
"architectures": [
|
4 |
+
"LlamaForCausalLM"
|
5 |
+
],
|
6 |
+
"attention_bias": false,
|
7 |
+
"attention_dropout": 0.0,
|
8 |
+
"bos_token_id": 128000,
|
9 |
+
"eos_token_id": [
|
10 |
+
128001,
|
11 |
+
128008,
|
12 |
+
128009
|
13 |
+
],
|
14 |
+
"freeze_mm_mlp_adapter": false,
|
15 |
+
"hidden_act": "silu",
|
16 |
+
"hidden_size": 4096,
|
17 |
+
"image_aspect_ratio": "square",
|
18 |
+
"initializer_range": 0.02,
|
19 |
+
"intermediate_size": 14336,
|
20 |
+
"max_position_embeddings": 131072,
|
21 |
+
"mlp_bias": false,
|
22 |
+
"mm_hidden_size": 3072,
|
23 |
+
"mm_patch_merge_type": "flat",
|
24 |
+
"mm_projector_lr": null,
|
25 |
+
"mm_projector_type": "mlp2x_gelu",
|
26 |
+
"mm_use_im_patch_token": false,
|
27 |
+
"mm_use_im_start_end": false,
|
28 |
+
"mm_vision_select_feature": "patch",
|
29 |
+
"mm_vision_select_layer": -2,
|
30 |
+
"mm_vision_tower": "/leonardo_scratch/large/userexternal/fcocchi0/rag_mlmm/hf_models/openai/clip-vit-large-patch14-336/models--openai--clip-vit-large-patch14-336/snapshots/ce19dc912ca5cd21c8a653c79e251e808ccabcd1",
|
31 |
+
"model_type": "llava_llama",
|
32 |
+
"num_attention_heads": 32,
|
33 |
+
"num_hidden_layers": 32,
|
34 |
+
"num_key_value_heads": 8,
|
35 |
+
"pretraining_tp": 1,
|
36 |
+
"rms_norm_eps": 1e-05,
|
37 |
+
"rope_scaling": {
|
38 |
+
"factor": 8.0,
|
39 |
+
"high_freq_factor": 4.0,
|
40 |
+
"low_freq_factor": 1.0,
|
41 |
+
"original_max_position_embeddings": 8192,
|
42 |
+
"rope_type": "llama3"
|
43 |
+
},
|
44 |
+
"rope_theta": 500000.0,
|
45 |
+
"s2": true,
|
46 |
+
"s2_scales": "336,672,1008",
|
47 |
+
"tie_word_embeddings": false,
|
48 |
+
"tokenizer_model_max_length": 4096,
|
49 |
+
"tokenizer_padding_side": "right",
|
50 |
+
"torch_dtype": "bfloat16",
|
51 |
+
"transformers_version": "4.43.1",
|
52 |
+
"tune_mm_mlp_adapter": true,
|
53 |
+
"use_cache": true,
|
54 |
+
"use_mm_proj": true,
|
55 |
+
"vocab_size": 128256
|
56 |
+
}
|
mm_projector.bin
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:3088d74e4d2c8038b83f69c43326c956b5765ebce7ddc59b1792a6cabb5bfd03
|
3 |
+
size 58738744
|
trainer_state.json
ADDED
The diff for this file is too large to render.
See raw diff
|
|