Commit
•
684e0c4
1
Parent(s):
4869712
Upload FlaxRPTForCausalLM
Browse files- config.json +66 -0
- flax_model.msgpack +3 -0
- generation_config.json +6 -0
config.json
ADDED
|
@@ -0,0 +1,66 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"add_null_attn": true,
|
| 3 |
+
"architectures": [
|
| 4 |
+
"RPTForCausalLM"
|
| 5 |
+
],
|
| 6 |
+
"attn_pdrop": 0.05,
|
| 7 |
+
"augment_across_neighbors": true,
|
| 8 |
+
"augment_neighbors": true,
|
| 9 |
+
"aux_loss_schedule_steps": 12500,
|
| 10 |
+
"aux_scale": 0.1,
|
| 11 |
+
"bos_token_id": 0,
|
| 12 |
+
"cca_freq": 1,
|
| 13 |
+
"chunk_size": 64,
|
| 14 |
+
"document_length": 16384,
|
| 15 |
+
"embd_pdrop": 0.0,
|
| 16 |
+
"eos_token_id": 1,
|
| 17 |
+
"fcm_max_ratio": 0.0,
|
| 18 |
+
"fcm_min_ratio": 0.0,
|
| 19 |
+
"gated_ff": true,
|
| 20 |
+
"hidden_size": 2048,
|
| 21 |
+
"initializer_range": 1,
|
| 22 |
+
"intermediate_size": 5504,
|
| 23 |
+
"margin_schedule_steps": 56250,
|
| 24 |
+
"max_margin": 4,
|
| 25 |
+
"max_sequence_length": 4096,
|
| 26 |
+
"mesh_dim": null,
|
| 27 |
+
"model_type": "rpt",
|
| 28 |
+
"mult_in_complex": false,
|
| 29 |
+
"n_windows": 1,
|
| 30 |
+
"num_attention_heads": 16,
|
| 31 |
+
"num_document_chunks": 256,
|
| 32 |
+
"num_hidden_layers": 22,
|
| 33 |
+
"num_key_value_heads": 16,
|
| 34 |
+
"num_neighbors": 2,
|
| 35 |
+
"num_scored_neighbors": 20,
|
| 36 |
+
"num_sequence_chunks": 64,
|
| 37 |
+
"palm_init": true,
|
| 38 |
+
"remat_attention": "",
|
| 39 |
+
"remat_block": "nothing_saveable",
|
| 40 |
+
"remat_mlp": "",
|
| 41 |
+
"resid_pdrop": 0.05,
|
| 42 |
+
"retriever_fill_value": -10000.0,
|
| 43 |
+
"return_ret_metrics": true,
|
| 44 |
+
"rms_norm_eps": 1e-06,
|
| 45 |
+
"rms_one_baseline": true,
|
| 46 |
+
"rot_dim": 0,
|
| 47 |
+
"run_modules": "all",
|
| 48 |
+
"scan_attention": false,
|
| 49 |
+
"scan_key_chunk_size": 2048,
|
| 50 |
+
"scan_mlp": false,
|
| 51 |
+
"scan_mlp_chunk_size": 1024,
|
| 52 |
+
"scan_query_chunk_size": 1024,
|
| 53 |
+
"scheduled_sampling_max_prob": 1.0,
|
| 54 |
+
"scheduled_sampling_min_prob": 0.01,
|
| 55 |
+
"sliding_window": false,
|
| 56 |
+
"ss_schedule_steps": 56250,
|
| 57 |
+
"stride": 1024,
|
| 58 |
+
"threshold_nei_scores": 0.0,
|
| 59 |
+
"tie_word_embeddings": true,
|
| 60 |
+
"transformers_version": "4.29.0.dev0",
|
| 61 |
+
"use_cache": true,
|
| 62 |
+
"use_cca_norm2": false,
|
| 63 |
+
"use_xnei_bias": true,
|
| 64 |
+
"vocab_size": 50277,
|
| 65 |
+
"window_length": 2048
|
| 66 |
+
}
|
flax_model.msgpack
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:6bd76d6fafe2dd73facb0e698f95585797be32325eb8619622401277fece5446
|
| 3 |
+
size 2952751330
|
generation_config.json
ADDED
|
@@ -0,0 +1,6 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"_from_model_config": true,
|
| 3 |
+
"bos_token_id": 0,
|
| 4 |
+
"eos_token_id": 1,
|
| 5 |
+
"transformers_version": "4.29.0.dev0"
|
| 6 |
+
}
|