git commited on
Commit
ff0d9fc
·
1 Parent(s): 0c292f3
This view is limited to 50 files because it contains too many changes.   See raw diff
.gitattributes CHANGED
@@ -33,3 +33,4 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ *.json filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,95 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ library_name: mlc-llm
3
+ tags:
4
+ - mlc-llm
5
+ - web-llm
6
+ language:
7
+ - en
8
+ base_model:
9
+ - Sao10K/L3-8B-Stheno-v3.2
10
+ pipeline_tag: text-generation
11
+ ---
12
+
13
+ This is a MLC converted weight from [L3-8B-Stheno-v3.2](https://huggingface.co/Sao10K/L3-8B-Stheno-v3.2) model in MLC format `q4f16_1`.
14
+
15
+ The model can be used for projects [MLC-LLM](https://github.com/mlc-ai/mlc-llm) and [WebLLM](https://github.com/mlc-ai/web-llm).
16
+
17
+ ---
18
+
19
+ *Just message me on discord if you want to host this privately for a service or something. We can talk.*
20
+
21
+ *Train used 1x H100 SXM for like a total of 24 Hours over multiple runs.*
22
+
23
+ Support me here if you're interested:
24
+ <br>Ko-fi: https://ko-fi.com/sao10k
25
+ <br> *wink* Euryale v2?
26
+
27
+ If not, that's fine too. Feedback would be nice.
28
+
29
+ Contact Me in Discord:
30
+ <br>`sao10k` // `Just ping me in the KoboldAI discord, I'll respond faster.`
31
+
32
+ `Art by navy_(navy.blue)` - [Danbooru](https://danbooru.donmai.us/posts/3214477)
33
+
34
+ ---
35
+
36
+ ![Stheno](https://huggingface.co/Sao10K/L3-8B-Stheno-v3.2/resolve/main/Stheno.png?)
37
+
38
+ Stheno-v3.2-Zeta
39
+
40
+ I have done a test run with multiple variations of the models, merged back to its base at various weights, different training runs too, and this Sixth iteration is the one I like most.
41
+
42
+
43
+ Changes compared to v3.1
44
+ <br>\- Included a mix of SFW and NSFW Storywriting Data, thanks to [Gryphe](https://huggingface.co/datasets/Gryphe/Opus-WritingPrompts)
45
+ <br>\- Included More Instruct / Assistant-Style Data
46
+ <br>\- Further cleaned up Roleplaying Samples from c2 Logs -> A few terrible, really bad samples escaped heavy filtering. Manual pass fixed it.
47
+ <br>\- Hyperparameter tinkering for training, resulting in lower loss levels.
48
+
49
+
50
+ Testing Notes - Compared to v3.1
51
+ <br>\- Handles SFW / NSFW seperately better. Not as overly excessive with NSFW now. Kinda balanced.
52
+ <br>\- Better at Storywriting / Narration.
53
+ <br>\- Better at Assistant-type Tasks.
54
+ <br>\- Better Multi-Turn Coherency -> Reduced Issues?
55
+ <br>\- Slightly less creative? A worthy tradeoff. Still creative.
56
+ <br>\- Better prompt / instruction adherence.
57
+
58
+ ---
59
+
60
+ **Recommended Samplers:**
61
+
62
+ ```
63
+ Temperature - 1.12-1.22
64
+ Min-P - 0.075
65
+ Top-K - 50
66
+ Repetition Penalty - 1.1
67
+ ```
68
+
69
+ **Stopping Strings:**
70
+
71
+ ```
72
+ \n\n{{User}} # Or Equivalent, depending on Frontend
73
+ <|eot_id|>
74
+ <|end_of_text|>
75
+ ```
76
+
77
+ **Prompting Template - Llama-3-Instruct**
78
+
79
+ ```
80
+ <|begin_of_text|><|start_header_id|>system<|end_header_id|>
81
+
82
+ {system_prompt}<|eot_id|><|start_header_id|>user<|end_header_id|>
83
+
84
+ {input}<|eot_id|><|start_header_id|>assistant<|end_header_id|>
85
+
86
+ {output}<|eot_id|>
87
+ ```
88
+
89
+ **Basic Roleplay System Prompt**
90
+ ```
91
+ You are an expert actor that can fully immerse yourself into any role given. You do not break character for any reason, even if someone tries addressing you as an AI or language model.
92
+ Currently your role is {{char}}, which is described in detail below. As {{char}}, continue the exchange with {{user}}.
93
+ ```
94
+
95
+ ---
mlc-chat-config.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cbf219d44f39830036af742f981e067ddd38eb17df5e4ac0b9dd03298e7a7a87
3
+ size 2150
ndarray-cache.json ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:38b017d5ad08c450fedb14d195895837bd163fd187cc2b681b392aaf4be8a34f
3
+ size 148403
params_shard_0.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9a516d9bdd583dbb7bf8cf7a72a58bc9f01cb49bd8afd79e0ac42047973212ed
3
+ size 262668288
params_shard_1.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a9c53809f46c0d422c99b7f53ed42af3445b1902ff5a26842e611aa506f3fe1
3
+ size 262668288
params_shard_10.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35771c79bb41376ba7e86a80cadb711f680fc4183660674088e51624b53116e0
3
+ size 29360128
params_shard_100.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ddf97652c1119364d47c7f98e0e0ecf945af56322df9e6fbf4ddc0b54d195684
3
+ size 33046528
params_shard_101.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4deb2c72aa6e27b68163d5098d8a030c473a4082b6d1d8b261eea91257b59060
3
+ size 29360128
params_shard_102.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a407d74406b27bfa17d950d89a26fab96aeeb3f4482abd6ff52de784389f3de
3
+ size 58720256
params_shard_103.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4fbb58816b9b9afcb10883ccd38c4853119385cefbbcc94b3247adc1efdc6c66
3
+ size 22036480
params_shard_104.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d2e429746835e2ba4327eb333e1a4a4a343463ad23c389c1b75a4426932edb2d
3
+ size 29360128
params_shard_105.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:26088e53f0369f5bdc32799b28407ca7943cae6980c8106f18fcb996424efd6e
3
+ size 58720256
params_shard_106.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9654a105f9dd021cb51dd13adced0b9fba695c2a5032d4bf07aa3b5815f63988
3
+ size 27271168
params_shard_107.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f5d33a0ad2f309f63bd08ff220ae8e0678db71f29a4bb27ff2361e160c599601
3
+ size 30949376
params_shard_11.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:99d9da280a703ea655c7f491931b9176825f59b93d54603ec9ff7448ab67e12b
3
+ size 58720256
params_shard_12.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b35a84bfb6d384a7405e2bf31cd3b2e5862487c59c4b2b9f273e4845feb12a4e
3
+ size 22036480
params_shard_13.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:95d162d4c07e7776894ffc3367d35c825a1be9e2d6bc29f0c038dc52843f943b
3
+ size 29360128
params_shard_14.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d5f5c6ce87c2b0a3ea2db39b935184d2c22777446888a7e5f06959e8b45c89ae
3
+ size 58720256
params_shard_15.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77001c60510462ed620f52df0acbfb5198d097ee492da89d635b3526daf4c60a
3
+ size 27271168
params_shard_16.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:89e0cd8ff9ca141eabcec563c30388f98ea7eedc3acf1d12a15bf6fb03122407
3
+ size 29360128
params_shard_17.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:91f3daeba5d11e9a6d80840cfbb7e3fb807316340b9f97c79853adb4a0adac3b
3
+ size 30949376
params_shard_18.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:004530b41f46e166877ef1e46434ff5a70b5adb6d9d066bd32cdd3cfac27e7e1
3
+ size 58720256
params_shard_19.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cab52c1316427b219ab3e786d6f968f7302cb3e4b62384d53725a4dc6e28f989
3
+ size 25174016
params_shard_2.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0aa09ea587f0f427f0017b2c46d1d7a846b1af389d95c074cae80aa4bbeed3ff
3
+ size 32833536
params_shard_20.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4caa38b171a161f7bd19abdd621d6de2879d5e03375f9aa17e4aa27c9cbd28cd
3
+ size 29360128
params_shard_21.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f8a4305c99c132c9016c305a5c6dff632ec5f2ae651b3eab4ad2e2124013d963
3
+ size 58720256
params_shard_22.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11ab4a4817145bb614a6f1df3bd10d10cd37291699910f75741014e94f6f3ad2
3
+ size 33046528
params_shard_23.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8d9487aec6ffc15b0e82aa8f31d1c4811d3ad65f7a58ddd02af70c89a2c8354d
3
+ size 29360128
params_shard_24.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3d2c380f5065c72970d8662249018cbcb4352bd5d823de1db5bd1dfb6f3d70d7
3
+ size 58720256
params_shard_25.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e68b3a1192b1b7bd5c3f5dd3d71088d360d99b0da8eb63218fdf18682126e06a
3
+ size 22036480
params_shard_26.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:29cf8ad249515db029392c70a1817fb8f2ac4eb2081f33f7ef2ebc8b70d4fe32
3
+ size 29360128
params_shard_27.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:77723fbb1aaa07163c252ab7847f6e266734555b21def784c90bafef962cac7e
3
+ size 58720256
params_shard_28.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:78d1f5fa3132ed910e27e00fed606aa19b03393d7c98b43e2ed5f20841e8c546
3
+ size 27271168
params_shard_29.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d0a3749533ecda9f907043e581781487cc33815143917f54045ec23723dbdd62
3
+ size 29360128
params_shard_3.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ce1bf94533eec237674ae1d66ccd0be9a2e35230c28295368ac01a029727300a
3
+ size 29360128
params_shard_30.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bef4fad728146a9b60236df2a42fc7a5caa92cc32f8588e3c02ea420d159db37
3
+ size 30949376
params_shard_31.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:be3493f6cb061b252b824f5d8b2185dea6199fa2a00efc125f5587a0dd9700a3
3
+ size 58720256
params_shard_32.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:34f772e89bcd580cad9983e28dfcf75debd071d2a73dde624bbd3933057bb060
3
+ size 25174016
params_shard_33.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ec799f6ffb90074407d0b1f298d3486c28b18c47a9d4e1f99e09f6d34825ea7
3
+ size 29360128
params_shard_34.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:32d0a2e1628acc69172bb89ce3f84eb9e602048849a28275e9efe122a2b94db1
3
+ size 58720256
params_shard_35.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6d3b755cb6d0a89d9d8f65f322a24b29060e9a2f74c44afac8da6a44e2c5957a
3
+ size 33046528
params_shard_36.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e324da2dfd72b5ca2093c78484b9da47dade58dcfb0400566609fbf3845a4e78
3
+ size 29360128
params_shard_37.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:29b5635a52a764c24d5fc25655699f2984500c899baa60073e7b83cd6fd364e6
3
+ size 58720256
params_shard_38.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:35a9294354ea746d5de4bfd25cca28abd4a0d2f43c4fd73446036f1dae2a37d6
3
+ size 22036480
params_shard_39.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2ff17a71a514a4fd7804bd4d7af0e3758fafa849ba73bec0ad3a7b24d385f197
3
+ size 29360128
params_shard_4.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52e0a818341792cad69c3124e8ed39e41448b42caf2c9308a5bf78fd4dc18a43
3
+ size 32841728
params_shard_40.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d86f45d8ca3648fd0e5e010a97d1873feed4e7812dfcab277a7e3be324798666
3
+ size 58720256
params_shard_41.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b30f229a39e278f9b9572a23470ebba93e1e8f6fbd5304367452ddbd0488146f
3
+ size 27271168
params_shard_42.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1c25e860c644ccc91d894338cab89313f0a8382e9f82ecd4134ec458f496932b
3
+ size 29360128