Hugging Face's logo

LoganDark
/
rwkv-4-raven-ggml

Model card Files Community
LoganDark commited on
Commit
d1d5e64
•
0 Parent(s):

initial commit

Browse files
Files changed (27) hide show
  1. .gitattributes +1 -0
  2. README.md +27 -0
  3. RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-Q4_0.bin +3 -0
  4. RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-Q4_2.bin +3 -0
  5. RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-Q5_1.bin +3 -0
  6. RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-Q8_0.bin +3 -0
  7. RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-f16.bin +3 -0
  8. RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-f32-00.bin +3 -0
  9. RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-f32-01.bin +3 -0
  10. RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-Q4_0.bin +3 -0
  11. RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-Q4_2.bin +3 -0
  12. RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-Q5_1.bin +3 -0
  13. RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-Q8_0.bin +3 -0
  14. RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-f16.bin +3 -0
  15. RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-f32.bin +3 -0
  16. RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-Q4_0.bin +3 -0
  17. RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-Q4_2.bin +3 -0
  18. RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-Q5_1.bin +3 -0
  19. RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-Q8_0.bin +3 -0
  20. RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-f16.bin +3 -0
  21. RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-f32.bin +3 -0
  22. RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-Q4_0.bin +3 -0
  23. RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-Q4_2.bin +3 -0
  24. RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-Q5_1.bin +3 -0
  25. RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-Q8_0.bin +3 -0
  26. RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-f16.bin +3 -0
  27. RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-f32.bin +3 -0
.gitattributes ADDED
@@ -0,0 +1 @@
 
 
1
+ *.bin filter=lfs diff=lfs merge=lfs -text
README.md ADDED
@@ -0,0 +1,27 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: apache-2.0
3
+ ---
4
+
5
+ # rwkv-4-raven-ggml
6
+
7
+ GGML models converted from the [`rwkv-4-raven`](https://huggingface.co/BlinkDL/rwkv-4-raven) checkpoints, for use with [`rwkv.cpp`](https://github.com/saharNooby/rwkv.cpp).
8
+
9
+ These models retain the original models' license (Apache 2.0).
10
+
11
+ ## Available models
12
+
13
+ | Name | `f32` | `f16` | `Q4_0` | `Q4_1` | `Q4_2` | `Q5_1` | `Q8_0` |
14
+ | ---------------------------------------------- | ----- | ----- | ------ | ------ | ------ | ------ | ------ |
15
+ | `RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096` | Yes | Yes | Yes | No | Yes | Yes | Yes |
16
+ | `RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096` | Yes | Yes | Yes | No | Yes | Yes | Yes |
17
+ | `RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192` | Yes | Yes | Yes | No | Yes | Yes | Yes |
18
+ | `RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192` | Split | Yes | Yes | No | Yes | Yes | Yes |
19
+
20
+ - The original PyTorch checkpoints (`.pth`) can be downloaded from the [`rwkv-4-raven`](https://huggingface.co/BlinkDL/rwkv-4-raven) repository.
21
+ - All `f32` and `f16` models were converted directly from the PyTorch checkpoints using [rwkv.cpp `convert_pytorch_to_ggml.py`](https://github.com/saharNooby/rwkv.cpp/blob/1c363e6d5f4ec7817ceffeeb17bd972b1ce9d9d0/rwkv/convert_pytorch_to_ggml.py).
22
+ - All quantized models were converted directly from their respective `f32` version using [rwkv.cpp `quantize.py`](https://github.com/saharNooby/rwkv.cpp/blob/1c363e6d5f4ec7817ceffeeb17bd972b1ce9d9d0/rwkv/quantize.py). Quantized models were never converted from other quantized models.
23
+ - Conversion and quantization took about an hour, and running `git add` on this repository took another hour. (lol)
24
+ Total time spent was about 2 hours.
25
+
26
+ - The `f32` version of RWKV-4-Raven-14B is too large to upload to HuggingFace (>50GB), so it has been split into two files. These files must be exactly concatenated together (using a utility like `cat`) to result in the original `.bin`.
27
+ - `Q4_1` is not offered because it is an awkward medium between `Q4_0` (fastest speed) and `Q4_2` (lowest size)
RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-Q4_0.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ada8296fc3bc1c732a1cc4a2134940627e552bda22d14613cf9b23e0bb54305a
3
+ size 10588149029
RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-Q4_2.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:0d1eb7eda3cba404602f02b6d35f6f3a8932c56fed2f908d0ebd75b3d3ca16ad
3
+ size 10588149029
RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-Q5_1.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:558b73049af42b4b4ffe0885d78b105f1ff58ef36abba182842b22e88f10dabb
3
+ size 12292085029
RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-Q8_0.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11b0bfa6ed67c9ad36c1b1e4e8ac0f002776f6efa5bfa700c52985e134c13c1e
3
+ size 17403893029
RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-f16.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f25f80555c840ad42b19411ca50788632eae9a714028b5d2431f6525d3296bce
3
+ size 28301772069
RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-f32-00.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a5d185b1fd12e20c1901ebd536a952468787cd3f33c9d304c48512c8000bcc1c
3
+ size 50000000000
RWKV-4-Raven-14B-v11x-Eng99-20230501-ctx8192-f32-01.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:52a81fc29d51fb031fd26b5185c485c2a9ecaf81ef612f3c354381bc8c3b7aa9
3
+ size 6594421029
RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-Q4_0.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fe66a99c9c979676cdde7aceacb4042af09c2c0eee563c6717bc7d58a7f36d7e
3
+ size 1643841109
RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-Q4_2.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8314a911e49cce4f679e1dd84dd0dc75d0f516f99500bd4f2839b69d75ef6326
3
+ size 1643841109
RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-Q5_1.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:e946c020632b14e01f80188ceb7039d55c955aedcf776e5f27c340f5aea35914
3
+ size 1807418965
RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-Q8_0.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b42779bedba11b182e4ddd72224e6133515c8b10ffd52fd8d074d96446822808
3
+ size 2298152533
RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-f16.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:098c6ea8368f68317c99283195651685bdaac417857a21e447eadced2e62f8eb
3
+ size 3031328341
RWKV-4-Raven-1B5-v11-Eng99-20230425-ctx4096-f32.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1a996cf24925f5e79aeef69d750983727b0eda60fcc8c318edd15307196fd51c
3
+ size 6060443221
RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-Q4_0.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:718c4ac41979c0b5e4b96ec9cbca340949c1500b63bc7b5bf10599a559308b09
3
+ size 2737278397
RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-Q4_2.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c2b4f8ea05bba27bb8597f8c797990c37c50ecb22bb673af2314fbfc7b19a88a
3
+ size 2737278397
RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-Q5_1.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:00f446633c48cfbdd47da5789d76a96ae96389f031ad1e27220051f07ed0ac27
3
+ size 3078065597
RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-Q8_0.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:a9535514d0c396f254a06169152fe0edfe7c5cadaecea41be1e46230f8240324
3
+ size 4100427197
RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-f16.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:20c1cbb2fce41a0e045d866ebb6766153c4f62cc8fee89091b61eae6603b6969
3
+ size 5971101117
RWKV-4-Raven-3B-v11-Eng99-20230425-ctx4096-f32.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d65a036d0bac85bd5064532390bd97246ffca80392e9667451371a99258647a6
3
+ size 11938532797
RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-Q4_0.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:55827e15d1650f0f26e16ed6bb60867c02f284ebdeedcd0f24a52db80d8a4e18
3
+ size 6015409597
RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-Q4_2.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:810d0bf9af43105f3b936dea78901c5e83e76a7ad3ec75d53b66a2c5a246742a
3
+ size 6015409597
RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-Q5_1.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:92e124d54855c4ab0b64f2002cdb8e3b5f0a9ad853f666e127b96b037eb5cae4
3
+ size 6887824829
RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-Q8_0.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8a630d00f85682f837fbff61f3db4cdcf46461d4f75c66d2d8b05f70ebfd11ce
3
+ size 9505070525
RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-f16.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e6886514a5c1162b760ae6ce873e71b8934d7e457b680b8cd7dcbfc4e118c8f
3
+ size 14788238781
RWKV-4-Raven-7B-v11x-Eng99-20230429-ctx8192-f32.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2814125c464516b846fa0c677a6562f26a82a0a9727f289c9573a55fedb1057b
3
+ size 29570620861