Upload chrono-2016-v3-ft weights from jjjlimaus/chrono-2016-v3-ft
Browse files- README.md +12 -0
- config.json +72 -0
- generation_config.json +10 -0
- model.safetensors +3 -0
- tokenizer.json +0 -0
- tokenizer_config.json +8 -0
README.md
ADDED
|
@@ -0,0 +1,12 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: apache-2.0
|
| 3 |
+
tags:
|
| 4 |
+
- text-generation
|
| 5 |
+
- pytorch
|
| 6 |
+
---
|
| 7 |
+
|
| 8 |
+
# lm-2b
|
| 9 |
+
|
| 10 |
+
A 2B-parameter causal language model (PyTorch / safetensors).
|
| 11 |
+
|
| 12 |
+
Checkpoint step: 3600
|
config.json
ADDED
|
@@ -0,0 +1,72 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"layer_types": [
|
| 3 |
+
"sliding_attention",
|
| 4 |
+
"sliding_attention",
|
| 5 |
+
"sliding_attention",
|
| 6 |
+
"full_attention",
|
| 7 |
+
"sliding_attention",
|
| 8 |
+
"sliding_attention",
|
| 9 |
+
"sliding_attention",
|
| 10 |
+
"full_attention",
|
| 11 |
+
"sliding_attention",
|
| 12 |
+
"sliding_attention",
|
| 13 |
+
"sliding_attention",
|
| 14 |
+
"full_attention",
|
| 15 |
+
"sliding_attention",
|
| 16 |
+
"sliding_attention",
|
| 17 |
+
"sliding_attention",
|
| 18 |
+
"full_attention",
|
| 19 |
+
"sliding_attention",
|
| 20 |
+
"sliding_attention",
|
| 21 |
+
"sliding_attention",
|
| 22 |
+
"full_attention",
|
| 23 |
+
"sliding_attention",
|
| 24 |
+
"sliding_attention",
|
| 25 |
+
"sliding_attention",
|
| 26 |
+
"full_attention",
|
| 27 |
+
"sliding_attention",
|
| 28 |
+
"sliding_attention",
|
| 29 |
+
"sliding_attention",
|
| 30 |
+
"full_attention"
|
| 31 |
+
],
|
| 32 |
+
"aux_gate_dim": 12,
|
| 33 |
+
"architectures": [
|
| 34 |
+
"NanochronoForCausalLM"
|
| 35 |
+
],
|
| 36 |
+
"num_attention_heads": 14,
|
| 37 |
+
"mix_gate_dim": 24,
|
| 38 |
+
"tap_layer": 14,
|
| 39 |
+
"num_hidden_layers": 28,
|
| 40 |
+
"num_key_value_heads": 14,
|
| 41 |
+
"bos_token_id": 32759,
|
| 42 |
+
"model_type": "sn38-nanochrono",
|
| 43 |
+
"hidden_size": 1792,
|
| 44 |
+
"final_logit_softcapping": 15.0,
|
| 45 |
+
"aux_layers": [
|
| 46 |
+
1,
|
| 47 |
+
3,
|
| 48 |
+
5,
|
| 49 |
+
7,
|
| 50 |
+
9,
|
| 51 |
+
11,
|
| 52 |
+
13,
|
| 53 |
+
15,
|
| 54 |
+
17,
|
| 55 |
+
19,
|
| 56 |
+
21,
|
| 57 |
+
23,
|
| 58 |
+
25,
|
| 59 |
+
27
|
| 60 |
+
],
|
| 61 |
+
"vocab_size": 32768,
|
| 62 |
+
"sliding_window": 512,
|
| 63 |
+
"intermediate_size": 7168,
|
| 64 |
+
"rope_theta": 100000.0,
|
| 65 |
+
"pad_token_id": 32763,
|
| 66 |
+
"max_position_embeddings": 2048,
|
| 67 |
+
"tie_word_embeddings": false,
|
| 68 |
+
"eos_token_id": 32763,
|
| 69 |
+
"use_cache": false,
|
| 70 |
+
"attention_multiplier": 1.2,
|
| 71 |
+
"torch_dtype": "bfloat16"
|
| 72 |
+
}
|
generation_config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token_id": 32759,
|
| 3 |
+
"do_sample": true,
|
| 4 |
+
"eos_token_id": 32763,
|
| 5 |
+
"max_new_tokens": 256,
|
| 6 |
+
"pad_token_id": 32763,
|
| 7 |
+
"temperature": 0.8,
|
| 8 |
+
"top_k": 50,
|
| 9 |
+
"transformers_version": "5.14.1"
|
| 10 |
+
}
|
model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:3a9e5a5b2008350be99ed46667f32cf57225b1381164bd4e105df676b8a23340
|
| 3 |
+
size 4037045748
|
tokenizer.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|
tokenizer_config.json
ADDED
|
@@ -0,0 +1,8 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"bos_token": "<|bos|>",
|
| 3 |
+
"eos_token": "<|assistant_end|>",
|
| 4 |
+
"pad_token": "<|assistant_end|>",
|
| 5 |
+
"clean_up_tokenization_spaces": false,
|
| 6 |
+
"model_max_length": 2048,
|
| 7 |
+
"add_bos_token": true
|
| 8 |
+
}
|