bartowski/Megrez-3B-Instruct-GGUF
1
0
Fork 0
Code Issues Pull Requests Actions Projects Releases Wiki Activity

初始化项目，由ModelHub XC社区提供模型

Browse Source 
Model: bartowski/Megrez-3B-Instruct-GGUF
Source: Original Platform
This commit is contained in:
ModelHub XC
2026-08-06 05:11:12 +08:00
commit 618ed96943
26 changed files with 293 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

58
.gitattributes vendored Normal file
View File

@@ -0,0 +1,58 @@
*.7z filter=lfs diff=lfs merge=lfs -text
*.arrow filter=lfs diff=lfs merge=lfs -text
*.bin filter=lfs diff=lfs merge=lfs -text
*.bz2 filter=lfs diff=lfs merge=lfs -text
*.ckpt filter=lfs diff=lfs merge=lfs -text
*.ftz filter=lfs diff=lfs merge=lfs -text
*.gz filter=lfs diff=lfs merge=lfs -text
*.h5 filter=lfs diff=lfs merge=lfs -text
*.joblib filter=lfs diff=lfs merge=lfs -text
*.lfs.* filter=lfs diff=lfs merge=lfs -text
*.mlmodel filter=lfs diff=lfs merge=lfs -text
*.model filter=lfs diff=lfs merge=lfs -text
*.msgpack filter=lfs diff=lfs merge=lfs -text
*.npy filter=lfs diff=lfs merge=lfs -text
*.npz filter=lfs diff=lfs merge=lfs -text
*.onnx filter=lfs diff=lfs merge=lfs -text
*.ot filter=lfs diff=lfs merge=lfs -text
*.parquet filter=lfs diff=lfs merge=lfs -text
*.pb filter=lfs diff=lfs merge=lfs -text
*.pickle filter=lfs diff=lfs merge=lfs -text
*.pkl filter=lfs diff=lfs merge=lfs -text
*.pt filter=lfs diff=lfs merge=lfs -text
*.pth filter=lfs diff=lfs merge=lfs -text
*.rar filter=lfs diff=lfs merge=lfs -text
*.safetensors filter=lfs diff=lfs merge=lfs -text
saved_model/**/* filter=lfs diff=lfs merge=lfs -text
*.tar.* filter=lfs diff=lfs merge=lfs -text
*.tar filter=lfs diff=lfs merge=lfs -text
*.tflite filter=lfs diff=lfs merge=lfs -text
*.tgz filter=lfs diff=lfs merge=lfs -text
*.wasm filter=lfs diff=lfs merge=lfs -text
*.xz filter=lfs diff=lfs merge=lfs -text
*.zip filter=lfs diff=lfs merge=lfs -text
*.zst filter=lfs diff=lfs merge=lfs -text
*tfevents* filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q8_0.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q6_K_L.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q6_K.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q5_K_L.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q5_K_M.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q5_K_S.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q4_K_L.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q4_K_M.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q4_K_S.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q4_0.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-IQ4_NL.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-IQ4_XS.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q3_K_XL.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q3_K_L.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q3_K_M.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-IQ3_M.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q3_K_S.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-IQ3_XS.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q2_K_L.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-Q2_K.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-IQ2_M.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct-f16.gguf filter=lfs diff=lfs merge=lfs -text
Megrez-3B-Instruct.imatrix filter=lfs diff=lfs merge=lfs -text

3
Megrez-3B-Instruct-IQ2_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:5fbf643799cba37ed9cc715c5e938c5254c46bd959e7b1e0c455f0a86a4e76e0
size 1132975680

3
Megrez-3B-Instruct-IQ3_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:c1ee466f609548d8b0364f8244657d3f9a40c7c2aa09d12d7d399435719c9225
size 1425675840

3
Megrez-3B-Instruct-IQ3_XS.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:331dcdf2df29121d5a97aec101bf60b64e18733640e53f73f5e4161dfe4927a1
size 1334498880

3
Megrez-3B-Instruct-IQ4_NL.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:0784fdca1d4cdcf94d35ebfe9b1468072772bfae6facf746837735d2caee252e
size 1733039680

3
Megrez-3B-Instruct-IQ4_XS.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:c518bc71a950f4d3532ff62996c157a48f12b6ecba83b21b43557067be59f8a6
size 1653413440

3
Megrez-3B-Instruct-Q2_K.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:0a4d21df9043049eb1d216aaeb9fda1025f54ee93714489d386f538aab6a4aec
size 1210226240

3
Megrez-3B-Instruct-Q2_K_L.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:48922946f11063e4dd42f44aaad894a1ef83916c29c547aa53edfea074fbcb16
size 1517426240

3
Megrez-3B-Instruct-Q3_K_L.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:f88af448ab784c7a4151fda3ef0bafff5fa81da8f428f3417f7e2e2e6a8fb4ec
size 1600083520

3
Megrez-3B-Instruct-Q3_K_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:2b3862f640ea52cef841a8cca6db5a2b3bd8c11440555da15944eec7fd4d8498
size 1498912320

3
Megrez-3B-Instruct-Q3_K_S.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:7f83345085fd2f8d3befe7c9a0def0a001802ab2d95586802600b798af15ea37
size 1381111360

3
Megrez-3B-Instruct-Q3_K_XL.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:09fca8c7d7acedc57ed6d3b61cbfdca98e4f8e65a88c159dab310b5666a85054
size 1875334720

3
Megrez-3B-Instruct-Q4_0.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:559bb79d8e91672a6fb5d5142c093b285e6d4f871e4302790cae6fcac211da05
size 1731073600

3
Megrez-3B-Instruct-Q4_K_L.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:6b4bbc64e3d6c42cb404d8154a4d01d41613f0babf43482d3c168dc20c85258d
size 2042410560

3
Megrez-3B-Instruct-Q4_K_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:ecb927e99e16a9882d263c957a305db23934001dac91fb52860701552c86cb91
size 1808938560

3
Megrez-3B-Instruct-Q4_K_S.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:4df40f254866bbd6d826506953cb44f17bc4c1ea81ba6bedee000e4d8a1c96da
size 1736480320

3
Megrez-3B-Instruct-Q5_K_L.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:26b14957ac91f2e2112a17d50785be9aa29e66636c3a5ec983dd39d3a1da4c4f
size 2288170560

3
Megrez-3B-Instruct-Q5_K_M.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:32f02cce7e2204d68bc574948e01761ca03d3c1cc37bfbecfd617d9b89c6a0bf
size 2094020160

3
Megrez-3B-Instruct-Q5_K_S.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:57f7dee3fafacb5d61ad5d0e5329f378b7a2f65960b8eb8a2438a8537c3483b7
size 2051544640

3
Megrez-3B-Instruct-Q6_K.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:8427d11e40ee9c698460f196bf734c65cb2575189d3fb64c2ed0deae92301692
size 2396919360

3
Megrez-3B-Instruct-Q6_K_L.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:42ba2f5af3b56fb901d26bb8a1618eeef151df0de7e933a4fb559864f72662ca
size 2549290560

3
Megrez-3B-Instruct-Q8_0.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:cc41d24d1c43b613168f5035542d6a8284bfd280552b768f81f5f13a1ed278c1
size 3102905920

3
Megrez-3B-Instruct-f16.gguf Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:d829867730264d0c9629853dbb6a2cf921fa29f6bfb2ff706ba57c0b9f6e0c25
size 5835756832

3
Megrez-3B-Instruct.imatrix Normal file
View File

@@ -0,0 +1,3 @@
version https://git-lfs.github.com/spec/v1
oid sha256:2a0b84e6d936ad32015390d21471993b271e1c6e9d21bf4b4c4c515d63ae4db9
size 2891018

165
README.md Normal file
View File

@@ -0,0 +1,165 @@
---
quantized_by: bartowski
pipeline_tag: text-generation
language:
- en
- zh
license: apache-2.0
base_model: Infinigence/Megrez-3B-Instruct
---
## Llamacpp imatrix Quantizations of Megrez-3B-Instruct
Using <a href="https://github.com/ggerganov/llama.cpp/">llama.cpp</a> release <a href="https://github.com/ggerganov/llama.cpp/releases/tag/b4381">b4381</a> for quantization.
Original model: https://huggingface.co/Infinigence/Megrez-3B-Instruct
All quants made using imatrix option with dataset from [here](https://gist.github.com/bartowski1182/eb213dccb3571f863da82e99418f81e8)
Run them in [LM Studio](https://lmstudio.ai/)
## Prompt format
```
<|role_start|>system<|role_end|>{system_prompt}<|turn_end|><|role_start|>user<|role_end|>{prompt}<|turn_end|><|role_start|>assistant<|role_end|>
```
## Download a file (not the whole branch) from below:
| Filename | Quant type | File Size | Split | Description |
| -------- | ---------- | --------- | ----- | ----------- |
| [Megrez-3B-Instruct-f16.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-f16.gguf) | f16 | 5.84GB | false | Full F16 weights. |
| [Megrez-3B-Instruct-Q8_0.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q8_0.gguf) | Q8_0 | 3.10GB | false | Extremely high quality, generally unneeded but max available quant. |
| [Megrez-3B-Instruct-Q6_K_L.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q6_K_L.gguf) | Q6_K_L | 2.55GB | false | Uses Q8_0 for embed and output weights. Very high quality, near perfect, *recommended*. |
| [Megrez-3B-Instruct-Q6_K.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q6_K.gguf) | Q6_K | 2.40GB | false | Very high quality, near perfect, *recommended*. |
| [Megrez-3B-Instruct-Q5_K_L.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q5_K_L.gguf) | Q5_K_L | 2.29GB | false | Uses Q8_0 for embed and output weights. High quality, *recommended*. |
| [Megrez-3B-Instruct-Q5_K_M.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q5_K_M.gguf) | Q5_K_M | 2.09GB | false | High quality, *recommended*. |
| [Megrez-3B-Instruct-Q5_K_S.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q5_K_S.gguf) | Q5_K_S | 2.05GB | false | High quality, *recommended*. |
| [Megrez-3B-Instruct-Q4_K_L.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q4_K_L.gguf) | Q4_K_L | 2.04GB | false | Uses Q8_0 for embed and output weights. Good quality, *recommended*. |
| [Megrez-3B-Instruct-Q3_K_XL.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q3_K_XL.gguf) | Q3_K_XL | 1.88GB | false | Uses Q8_0 for embed and output weights. Lower quality but usable, good for low RAM availability. |
| [Megrez-3B-Instruct-Q4_K_M.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q4_K_M.gguf) | Q4_K_M | 1.81GB | false | Good quality, default size for most use cases, *recommended*. |
| [Megrez-3B-Instruct-Q4_K_S.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q4_K_S.gguf) | Q4_K_S | 1.74GB | false | Slightly lower quality with more space savings, *recommended*. |
| [Megrez-3B-Instruct-Q4_0.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q4_0.gguf) | Q4_0 | 1.73GB | false | Legacy format, offers online repacking for ARM and AVX CPU inference. |
| [Megrez-3B-Instruct-IQ4_NL.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-IQ4_NL.gguf) | IQ4_NL | 1.73GB | false | Similar to IQ4_XS, but slightly larger. Offers online repacking for ARM CPU inference. |
| [Megrez-3B-Instruct-IQ4_XS.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-IQ4_XS.gguf) | IQ4_XS | 1.65GB | false | Decent quality, smaller than Q4_K_S with similar performance, *recommended*. |
| [Megrez-3B-Instruct-Q3_K_L.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q3_K_L.gguf) | Q3_K_L | 1.60GB | false | Lower quality but usable, good for low RAM availability. |
| [Megrez-3B-Instruct-Q2_K_L.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q2_K_L.gguf) | Q2_K_L | 1.52GB | false | Uses Q8_0 for embed and output weights. Very low quality but surprisingly usable. |
| [Megrez-3B-Instruct-Q3_K_M.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q3_K_M.gguf) | Q3_K_M | 1.50GB | false | Low quality. |
| [Megrez-3B-Instruct-IQ3_M.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-IQ3_M.gguf) | IQ3_M | 1.43GB | false | Medium-low quality, new method with decent performance comparable to Q3_K_M. |
| [Megrez-3B-Instruct-Q3_K_S.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q3_K_S.gguf) | Q3_K_S | 1.38GB | false | Low quality, not recommended. |
| [Megrez-3B-Instruct-IQ3_XS.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-IQ3_XS.gguf) | IQ3_XS | 1.33GB | false | Lower quality, new method with decent performance, slightly better than Q3_K_S. |
| [Megrez-3B-Instruct-Q2_K.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-Q2_K.gguf) | Q2_K | 1.21GB | false | Very low quality but surprisingly usable. |
| [Megrez-3B-Instruct-IQ2_M.gguf](https://huggingface.co/bartowski/Megrez-3B-Instruct-GGUF/blob/main/Megrez-3B-Instruct-IQ2_M.gguf) | IQ2_M | 1.13GB | false | Relatively low quality, uses SOTA techniques to be surprisingly usable. |
## Embed/output weights
Some of these quants (Q3_K_XL, Q4_K_L etc) are the standard quantization method with the embeddings and output weights quantized to Q8_0 instead of what they would normally default to.
## Downloading using huggingface-cli
<details>
<summary>Click to view download instructions</summary>
First, make sure you have hugginface-cli installed:
```
pip install -U "huggingface_hub[cli]"
```
Then, you can target the specific file you want:
```
huggingface-cli download bartowski/Megrez-3B-Instruct-GGUF --include "Megrez-3B-Instruct-Q4_K_M.gguf" --local-dir ./
```
If the model is bigger than 50GB, it will have been split into multiple files. In order to download them all to a local folder, run:
```
huggingface-cli download bartowski/Megrez-3B-Instruct-GGUF --include "Megrez-3B-Instruct-Q8_0/*" --local-dir ./
```
You can either specify a new local-dir (Megrez-3B-Instruct-Q8_0) or download them all in place (./)
</details>
## ARM/AVX information
Previously, you would download Q4_0_4_4/4_8/8_8, and these would have their weights interleaved in memory in order to improve performance on ARM and AVX machines by loading up more data in one pass.
Now, however, there is something called "online repacking" for weights. details in [this PR](https://github.com/ggerganov/llama.cpp/pull/9921). If you use Q4_0 and your hardware would benefit from repacking weights, it will do it automatically on the fly.
As of llama.cpp build [b4282](https://github.com/ggerganov/llama.cpp/releases/tag/b4282) you will not be able to run the Q4_0_X_X files and will instead need to use Q4_0.
Additionally, if you want to get slightly better quality for , you can use IQ4_NL thanks to [this PR](https://github.com/ggerganov/llama.cpp/pull/10541) which will also repack the weights for ARM, though only the 4_4 for now. The loading time may be slower but it will result in an overall speed incrase.
<details>
<summary>Click to view Q4_0_X_X information (deprecated</summary>
I'm keeping this section to show the potential theoretical uplift in performance from using the Q4_0 with online repacking.
<details>
<summary>Click to view benchmarks on an AVX2 system (EPYC7702)</summary>
| model | size | params | backend | threads | test | t/s | % (vs Q4_0) |
| ------------------------------ | ---------: | ---------: | ---------- | ------: | ------------: | -------------------: |-------------: |
| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | pp512 | 204.03 ± 1.03 | 100% |
| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | pp1024 | 282.92 ± 0.19 | 100% |
| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | pp2048 | 259.49 ± 0.44 | 100% |
| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | tg128 | 39.12 ± 0.27 | 100% |
| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | tg256 | 39.31 ± 0.69 | 100% |
| qwen2 3B Q4_0 | 1.70 GiB | 3.09 B | CPU | 64 | tg512 | 40.52 ± 0.03 | 100% |
| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | pp512 | 301.02 ± 1.74 | 147% |
| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | pp1024 | 287.23 ± 0.20 | 101% |
| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | pp2048 | 262.77 ± 1.81 | 101% |
| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | tg128 | 18.80 ± 0.99 | 48% |
| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | tg256 | 24.46 ± 3.04 | 83% |
| qwen2 3B Q4_K_M | 1.79 GiB | 3.09 B | CPU | 64 | tg512 | 36.32 ± 3.59 | 90% |
| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | pp512 | 271.71 ± 3.53 | 133% |
| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | pp1024 | 279.86 ± 45.63 | 100% |
| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | pp2048 | 320.77 ± 5.00 | 124% |
| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | tg128 | 43.51 ± 0.05 | 111% |
| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | tg256 | 43.35 ± 0.09 | 110% |
| qwen2 3B Q4_0_8_8 | 1.69 GiB | 3.09 B | CPU | 64 | tg512 | 42.60 ± 0.31 | 105% |
Q4_0_8_8 offers a nice bump to prompt processing and a small bump to text generation
</details>
</details>
## Which file should I choose?
<details>
<summary>Click here for details</summary>
A great write up with charts showing various performances is provided by Artefact2 [here](https://gist.github.com/Artefact2/b5f810600771265fc1e39442288e8ec9)
The first thing to figure out is how big a model you can run. To do this, you'll need to figure out how much RAM and/or VRAM you have.
If you want your model running as FAST as possible, you'll want to fit the whole thing on your GPU's VRAM. Aim for a quant with a file size 1-2GB smaller than your GPU's total VRAM.
If you want the absolute maximum quality, add both your system RAM and your GPU's VRAM together, then similarly grab a quant with a file size 1-2GB Smaller than that total.
Next, you'll need to decide if you want to use an 'I-quant' or a 'K-quant'.
If you don't want to think too much, grab one of the K-quants. These are in format 'QX_K_X', like Q5_K_M.
If you want to get more into the weeds, you can check out this extremely useful feature chart:
[llama.cpp feature matrix](https://github.com/ggerganov/llama.cpp/wiki/Feature-matrix)
But basically, if you're aiming for below Q4, and you're running cuBLAS (Nvidia) or rocBLAS (AMD), you should look towards the I-quants. These are in format IQX_X, like IQ3_M. These are newer and offer better performance for their size.
These I-quants can also be used on CPU and Apple Metal, but will be slower than their K-quant equivalent, so speed vs performance is a tradeoff you'll have to decide.
The I-quants are *not* compatible with Vulcan, which is also AMD, so if you have an AMD card double check if you're using the rocBLAS build or the Vulcan build. At the time of writing this, LM Studio has a preview with ROCm support, and other inference engines have specific builds for ROCm.
</details>
## Credits
Thank you kalomaze and Dampf for assistance in creating the imatrix calibration dataset.
Thank you ZeroWw for the inspiration to experiment with embed/output.
Want to support my work? Visit my ko-fi page here: https://ko-fi.com/bartowski

1
configuration.json Normal file
View File

@@ -0,0 +1 @@
{"framework": "pytorch", "task": "text-generation", "allow_remote": true}