iohadrubin
commited on
Commit
•
655eb8d
1
Parent(s):
e02e031
Upload FlaxGPTNeoXForCausalLM
Browse files- config.json +71 -0
- generation_config.json +6 -0
- model.safetensors +3 -0
config.json
ADDED
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"a_init_nei": null,
|
3 |
+
"a_init_query": null,
|
4 |
+
"append_next_chunk": true,
|
5 |
+
"apply_gating": true,
|
6 |
+
"apply_query_to_nei_att": false,
|
7 |
+
"apply_refactor": false,
|
8 |
+
"apply_tanh_in_cca": false,
|
9 |
+
"architectures": [
|
10 |
+
"GPTNeoXForCausalLM"
|
11 |
+
],
|
12 |
+
"atsc_margin_min": 1.0,
|
13 |
+
"attention_bias": true,
|
14 |
+
"attention_dropout": 0.0,
|
15 |
+
"aux_scale": 1.0,
|
16 |
+
"bos_token_id": 0,
|
17 |
+
"cca_freq": 3,
|
18 |
+
"cca_layernorm_init_scale": 1,
|
19 |
+
"chunk_size": 64,
|
20 |
+
"classifier_dropout": 0.1,
|
21 |
+
"do_last_gate": false,
|
22 |
+
"document_length": 16384,
|
23 |
+
"eos_token_id": 0,
|
24 |
+
"gate_num_blocks": 8,
|
25 |
+
"gate_refactor": false,
|
26 |
+
"hidden_act": "gelu",
|
27 |
+
"hidden_dropout": 0.0,
|
28 |
+
"hidden_size": 2048,
|
29 |
+
"initializer_range": 0.02,
|
30 |
+
"intermediate_size": 8192,
|
31 |
+
"layer_norm_eps": 1e-05,
|
32 |
+
"log_debug_metrics": false,
|
33 |
+
"lowres_ss": false,
|
34 |
+
"max_position_embeddings": 2048,
|
35 |
+
"max_sequence_length": 2048,
|
36 |
+
"model_type": "gpt_neox",
|
37 |
+
"mult_by_ndcg": true,
|
38 |
+
"n_query_aug_layers": null,
|
39 |
+
"null_attn_init": 0.0001,
|
40 |
+
"null_k_init_mult": 1000,
|
41 |
+
"num_attention_heads": 16,
|
42 |
+
"num_hidden_layers": 24,
|
43 |
+
"num_neighbors": 2,
|
44 |
+
"num_scored_neighbors": 20,
|
45 |
+
"pooling_size": 128,
|
46 |
+
"qk_layernorm": true,
|
47 |
+
"query_aug_dim": 128,
|
48 |
+
"remat_attention": "",
|
49 |
+
"ret_score_in_cca": false,
|
50 |
+
"retriever_fill_value": -10000.0,
|
51 |
+
"rope_scaling": null,
|
52 |
+
"rotary_emb_base": 10000,
|
53 |
+
"rotary_pct": 0.25,
|
54 |
+
"scheduled_sampling_max_prob": 1.0,
|
55 |
+
"scheduled_sampling_min_prob": 0.01,
|
56 |
+
"score_temp": 0.05,
|
57 |
+
"set_pt_params": false,
|
58 |
+
"ss_schedule_steps": 900000,
|
59 |
+
"stop_grad_trick": false,
|
60 |
+
"tanh_causal_att": false,
|
61 |
+
"tanh_xatt": false,
|
62 |
+
"threshold_nei_scores": 0.0,
|
63 |
+
"tie_word_embeddings": false,
|
64 |
+
"torch_dtype": "float16",
|
65 |
+
"transformers_version": "4.37.0.dev0",
|
66 |
+
"use_allowed_tar_mask": false,
|
67 |
+
"use_cache": true,
|
68 |
+
"use_dec_cache": true,
|
69 |
+
"use_parallel_residual": true,
|
70 |
+
"vocab_size": 50304
|
71 |
+
}
|
generation_config.json
ADDED
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"_from_model_config": true,
|
3 |
+
"bos_token_id": 0,
|
4 |
+
"eos_token_id": 0,
|
5 |
+
"transformers_version": "4.37.0.dev0"
|
6 |
+
}
|
model.safetensors
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:885f45cae8750ac2e0844e406d9e0684c68ed021fb3adf536327fa82ea0693e6
|
3 |
+
size 3014083970
|