Update metadata for renamed artifacts and Windows universal build
Browse files
README.md
CHANGED
|
@@ -1,119 +1,120 @@
|
|
| 1 |
-
---
|
| 2 |
-
license: mit
|
| 3 |
-
library_name: llama.cpp
|
| 4 |
-
base_model: Diffuspace/Clover-35B-A3B-Preview
|
| 5 |
-
base_model_relation: quantized
|
| 6 |
-
language:
|
| 7 |
-
- en
|
| 8 |
-
tags:
|
| 9 |
-
- llama.cpp
|
| 10 |
-
- gguf
|
| 11 |
-
- clover
|
| 12 |
-
- clover_moe
|
| 13 |
-
- prebuilt
|
| 14 |
-
- runtime
|
| 15 |
-
- linux
|
| 16 |
-
- x86-64
|
| 17 |
-
- cpu
|
| 18 |
-
- cuda
|
| 19 |
-
- cuda12
|
| 20 |
-
- converter
|
| 21 |
-
- quantize
|
| 22 |
-
- imatrix
|
| 23 |
-
- diffusion
|
| 24 |
-
- block-diffusion
|
| 25 |
-
- moe
|
| 26 |
-
---
|
| 27 |
-
|
| 28 |
-
# llama.cpp with Clover support (clover_moe)
|
| 29 |
-
|
| 30 |
-
**Prebuilt Clover runtime by IsValorum**
|
| 31 |
-
|
| 32 |
-
<p align="left">
|
| 33 |
-
<a href="https://ko-fi.com/isvalorum"><img src="https://ko-fi.com/img/githubbutton_sm.svg" alt="Support on Ko-fi" height="38"></a>
|
| 34 |
-
</p>
|
| 35 |
-
|
| 36 |
-
> [!IMPORTANT]
|
| 37 |
-
> **Independently computed using dedicated cloud compute clusters, funded 100% out-of-pocket.**
|
| 38 |
-
>
|
| 39 |
-
> If you consider these releases to be of high quality and they help you in your day-to-day workflow, consider [supporting on Ko-fi](https://ko-fi.com/isvalorum). Every contribution directly sustains future cloud compute hours and helps support my university studies.
|
| 40 |
-
|
| 41 |
-
Prebuilt Linux x86_64 binaries for [Diffuspace/Clover-35B-A3B-Preview](https://huggingface.co/Diffuspace/Clover-35B-A3B-Preview), built from llama.cpp PR [#31306](https://github.com/ggml-org/llama.cpp/pull/31306) at commit `4838ffb0c8ade92f34d9157deea38d92758ab824`.
|
| 42 |
-
|
| 43 |
-
Mainline llama.cpp does not load `clover_moe`. This build adds the converter, the runtime graph and the `draft-self` speculative mode. The PR is not merged upstream; pin this repository for reproducibility.
|
| 44 |
-
|
| 45 |
-
## Clover artifacts for this runtime
|
| 46 |
-
|
| 47 |
-
| Artifact | Size | Link |
|
| 48 |
-
|---|---|---|
|
| 49 |
-
| BF16 source | 69.38 GB | [Clover-35B-A3B-Preview-BF16-GGUF](https://huggingface.co/IsValorum/Clover-35B-A3B-Preview-BF16-GGUF) |
|
| 50 |
-
| Q8_0 | 36.91 GB | [Clover-35B-A3B-Preview-Q8_0-GGUF](https://huggingface.co/IsValorum/Clover-35B-A3B-Preview-Q8_0-GGUF) |
|
| 51 |
-
| 1.25M I-Matrix | 383,689,280 bytes | [Clover-35B-A3B-Preview-VAL-1.25M-I-Matrix-GGUF](https://huggingface.co/IsValorum/Clover-35B-A3B-Preview-VAL-1.25M-I-Matrix-GGUF) |
|
| 52 |
-
| APEX-I NanoPlus V2.1 | 12.56 GB / 11.70 GiB | [Clover-35B-A3B-Preview-APEX-I-NanoPlus-V2.1-GGUF](https://huggingface.co/IsValorum/Clover-35B-A3B-Preview-APEX-I-NanoPlus-V2.1-GGUF) |
|
| 53 |
-
| APEX-I MiniPlus V2.1 | 14.75 GB / 13.74 GiB | [Clover-35B-A3B-Preview-APEX-I-MiniPlus-V2.1-GGUF](https://huggingface.co/IsValorum/Clover-35B-A3B-Preview-APEX-I-MiniPlus-V2.1-GGUF) |
|
| 54 |
-
|
| 55 |
-
## Files
|
| 56 |
-
|
| 57 |
-
| File | Description |
|
| 58 |
-
|---|---|
|
| 59 |
-
| `llama-clover-
|
| 60 |
-
| `llama-clover-
|
| 61 |
-
| `llama-clover-
|
| 62 |
-
| `
|
| 63 |
-
| `
|
| 64 |
-
| `
|
| 65 |
-
|
| 66 |
-
|
| 67 |
-
|
| 68 |
-
|
| 69 |
-
>
|
| 70 |
-
|
| 71 |
-
|
| 72 |
-
-
|
| 73 |
-
-
|
| 74 |
-
-
|
| 75 |
-
-
|
| 76 |
-
-
|
| 77 |
-
|
| 78 |
-
|
| 79 |
-
|
| 80 |
-
|
| 81 |
-
|
| 82 |
-
|
| 83 |
-
|
| 84 |
-
|
| 85 |
-
|
| 86 |
-
|
| 87 |
-
|
| 88 |
-
|
| 89 |
-
|
| 90 |
-
|
| 91 |
-
|
| 92 |
-
|
| 93 |
-
|
| 94 |
-
|
| 95 |
-
|
| 96 |
-
|
| 97 |
-
|
| 98 |
-
./bin/llama-
|
| 99 |
-
|
| 100 |
-
|
| 101 |
-
|
| 102 |
-
|
| 103 |
-
|
| 104 |
-
|
| 105 |
-
|
| 106 |
-
|
| 107 |
-
|
| 108 |
-
|
| 109 |
-
|
| 110 |
-
|
| 111 |
-
|
| 112 |
-
./bin/llama-
|
| 113 |
-
|
| 114 |
-
|
| 115 |
-
|
| 116 |
-
|
| 117 |
-
|
| 118 |
-
|
| 119 |
-
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
library_name: llama.cpp
|
| 4 |
+
base_model: Diffuspace/Clover-35B-A3B-Preview
|
| 5 |
+
base_model_relation: quantized
|
| 6 |
+
language:
|
| 7 |
+
- en
|
| 8 |
+
tags:
|
| 9 |
+
- llama.cpp
|
| 10 |
+
- gguf
|
| 11 |
+
- clover
|
| 12 |
+
- clover_moe
|
| 13 |
+
- prebuilt
|
| 14 |
+
- runtime
|
| 15 |
+
- linux
|
| 16 |
+
- x86-64
|
| 17 |
+
- cpu
|
| 18 |
+
- cuda
|
| 19 |
+
- cuda12
|
| 20 |
+
- converter
|
| 21 |
+
- quantize
|
| 22 |
+
- imatrix
|
| 23 |
+
- diffusion
|
| 24 |
+
- block-diffusion
|
| 25 |
+
- moe
|
| 26 |
+
---
|
| 27 |
+
|
| 28 |
+
# llama.cpp with Clover support (clover_moe)
|
| 29 |
+
|
| 30 |
+
**Prebuilt Clover runtime by IsValorum**
|
| 31 |
+
|
| 32 |
+
<p align="left">
|
| 33 |
+
<a href="https://ko-fi.com/isvalorum"><img src="https://ko-fi.com/img/githubbutton_sm.svg" alt="Support on Ko-fi" height="38"></a>
|
| 34 |
+
</p>
|
| 35 |
+
|
| 36 |
+
> [!IMPORTANT]
|
| 37 |
+
> **Independently computed using dedicated cloud compute clusters, funded 100% out-of-pocket.**
|
| 38 |
+
>
|
| 39 |
+
> If you consider these releases to be of high quality and they help you in your day-to-day workflow, consider [supporting on Ko-fi](https://ko-fi.com/isvalorum). Every contribution directly sustains future cloud compute hours and helps support my university studies.
|
| 40 |
+
|
| 41 |
+
Prebuilt Linux and Windows x86_64 binaries for [Diffuspace/Clover-35B-A3B-Preview](https://huggingface.co/Diffuspace/Clover-35B-A3B-Preview), built from llama.cpp PR [#31306](https://github.com/ggml-org/llama.cpp/pull/31306) at commit `4838ffb0c8ade92f34d9157deea38d92758ab824`.
|
| 42 |
+
|
| 43 |
+
Mainline llama.cpp does not load `clover_moe`. This build adds the converter, the runtime graph and the `draft-self` speculative mode. The PR is not merged upstream; pin this repository for reproducibility.
|
| 44 |
+
|
| 45 |
+
## Clover artifacts for this runtime
|
| 46 |
+
|
| 47 |
+
| Artifact | Size | Link |
|
| 48 |
+
|---|---|---|
|
| 49 |
+
| BF16 source | 69.38 GB | [Clover-35B-A3B-Preview-BF16-GGUF](https://huggingface.co/IsValorum/Clover-35B-A3B-Preview-BF16-GGUF) |
|
| 50 |
+
| Q8_0 | 36.91 GB | [Clover-35B-A3B-Preview-Q8_0-GGUF](https://huggingface.co/IsValorum/Clover-35B-A3B-Preview-Q8_0-GGUF) |
|
| 51 |
+
| 1.25M I-Matrix | 383,689,280 bytes | [Clover-35B-A3B-Preview-VAL-1.25M-I-Matrix-GGUF](https://huggingface.co/IsValorum/Clover-35B-A3B-Preview-VAL-1.25M-I-Matrix-GGUF) |
|
| 52 |
+
| APEX-I NanoPlus V2.1 | 12.56 GB / 11.70 GiB | [Clover-35B-A3B-Preview-APEX-I-NanoPlus-V2.1-GGUF](https://huggingface.co/IsValorum/Clover-35B-A3B-Preview-APEX-I-NanoPlus-V2.1-GGUF) |
|
| 53 |
+
| APEX-I MiniPlus V2.1 | 14.75 GB / 13.74 GiB | [Clover-35B-A3B-Preview-APEX-I-MiniPlus-V2.1-GGUF](https://huggingface.co/IsValorum/Clover-35B-A3B-Preview-APEX-I-MiniPlus-V2.1-GGUF) |
|
| 54 |
+
|
| 55 |
+
## Files
|
| 56 |
+
|
| 57 |
+
| File | Description |
|
| 58 |
+
|---|---|
|
| 59 |
+
| `llama-clover-4838ffb-linux-x86_64-cpu.tar.gz` | Linux CPU build, `GGML_NATIVE=OFF` |
|
| 60 |
+
| `llama-clover-4838ffb-linux-x86_64-nvidia-cuda12.tar.gz` | CUDA 12 build, archs `80;86;89;90` |
|
| 61 |
+
| `llama-clover-4838ffb-converter-python.tar.gz` | HF-to-GGUF converter, conversion code and gguf-py |
|
| 62 |
+
| `llama-clover-4838ffb-windows-x86_64-nvidia-amd-universal-cuda12.8-vulkan-cpu.zip` | Windows x86_64 universal: Nvidia CUDA 12.8 sm_50-120, AMD/Intel Vulkan, any CPU. Tested with NanoPlus: CPU, CUDA -ngl 8, Vulkan Intel |
|
| 63 |
+
| `BUILD_INFO.json` | Compiler, flags and SHA256 per archive |
|
| 64 |
+
| `SHA256SUMS` | Checksums |
|
| 65 |
+
| `config.json` | Machine-readable build descriptor used by Hub download tracking |
|
| 66 |
+
|
| 67 |
+
## Compatibility
|
| 68 |
+
|
| 69 |
+
> [!NOTE]
|
| 70 |
+
> **Platform availability.** The precompiled builds published today run on Linux only. Universal builds are in progress for Windows, AMD GPUs, NVIDIA GPUs and any CPU; until they ship, any other platform needs a build from the same branch.
|
| 71 |
+
|
| 72 |
+
- Model: `Diffuspace/Clover-35B-A3B-Preview` (`clover_moe`). Works with the BF16 and Q8_0 GGUFs and with third-party `clover_moe` GGUFs such as the IQ4_XS. It does not load the original safetensors.
|
| 73 |
+
- Linux x86_64: the CPU package runs on any Linux; the CUDA package requires an NVIDIA GPU and a CUDA 12.x runtime, with archs 80/86/89/90 (A100, H100/H200, L40S, RTX 4090, A10G). Blackwell 100/120 is not in this build.
|
| 74 |
+
- glibc: the Linux x64 archives require `GLIBC_2.38` or newer. Ubuntu 24.04 or equivalent is the safe baseline; older distributions need their own build from the branch.
|
| 75 |
+
- Windows: no native binary here. Use WSL2 with the Linux CPU or CUDA package; a native Windows build must be compiled from the PR with CMake and MSVC. The GGUF files themselves are OS-agnostic.
|
| 76 |
+
- macOS: no binary here; use a Metal build of the PR or the third-party IQ4_XS Mac build.
|
| 77 |
+
- Frontends: not LM Studio or Ollama. Clover needs this `clover_moe` runtime until the PR merges upstream.
|
| 78 |
+
|
| 79 |
+
## Usage
|
| 80 |
+
|
| 81 |
+
Both runtime archives are dynamically linked and ship `llama-cli`, `llama-server`, `llama-diffusion-cli`, `llama-quantize`, `llama-imatrix`, and `llama-perplexity`. Point the loader at the bundled libraries before running.
|
| 82 |
+
|
| 83 |
+
CPU:
|
| 84 |
+
|
| 85 |
+
```bash
|
| 86 |
+
tar -xzf llama-clover-b4838ffb-linux-x64-cpu.tar.gz
|
| 87 |
+
cd llama-clover-b4838ffb-linux-x64-cpu
|
| 88 |
+
export LD_LIBRARY_PATH=$PWD/bin:$LD_LIBRARY_PATH
|
| 89 |
+
./bin/llama-cli -m clover-bf16.gguf -c 4096 -p "Hello"
|
| 90 |
+
```
|
| 91 |
+
|
| 92 |
+
CUDA 12:
|
| 93 |
+
|
| 94 |
+
```bash
|
| 95 |
+
tar -xzf llama-clover-b4838ffb-linux-x64-cuda12.tar.gz
|
| 96 |
+
cd llama-clover-b4838ffb-linux-x64-cuda12
|
| 97 |
+
export LD_LIBRARY_PATH=$PWD/bin:$LD_LIBRARY_PATH
|
| 98 |
+
./bin/llama-cli -m clover-bf16.gguf -ngl 99 -fa on -c 4096 -p "Hello"
|
| 99 |
+
./bin/llama-speculative-simple -m clover-bf16.gguf --spec-type draft-self --spec-draft-n-max 15 -ctk bf16 -ctv bf16 -n 128 -c 2048 -b 256 -ub 256 -ngl 99
|
| 100 |
+
```
|
| 101 |
+
|
| 102 |
+
## Conversion
|
| 103 |
+
|
| 104 |
+
```bash
|
| 105 |
+
tar -xzf llama-clover-b4838ffb-converter.tar.gz
|
| 106 |
+
python llama-clover-converter/llama.cpp/convert_hf_to_gguf.py Diffuspace/Clover-35B-A3B-Preview --outtype bf16 --outfile clover-bf16.gguf
|
| 107 |
+
```
|
| 108 |
+
|
| 109 |
+
## Quantization and importance matrix
|
| 110 |
+
|
| 111 |
+
```bash
|
| 112 |
+
./bin/llama-quantize --allow-requantize clover-bf16.gguf clover-q8_0.gguf Q8_0
|
| 113 |
+
./bin/llama-imatrix -m clover-q8_0.gguf -f calibration.raw -o clover.imatrix.gguf --output-format gguf --process-output --no-ppl --no-escape -c 2048 -b 2048 -ub 2048 --chunks 611 -ngl 99
|
| 114 |
+
```
|
| 115 |
+
|
| 116 |
+
The APEX-I V2.1 releases were produced with this build: `llama-quantize --imatrix imatrix-1.25M.gguf --tensor-type-file <recipe> clover-bf16.gguf <output> Q4_0 64`. The recipes carry 853 lines, including the 120 block-diffusion tensors that must stay in `F32`.
|
| 117 |
+
|
| 118 |
+
## Notes
|
| 119 |
+
|
| 120 |
+
The CUDA package requires a compatible NVIDIA driver and CUDA 12.x runtime. llama.cpp is MIT licensed; the Clover support is PR #31306 by Maphist0; the Clover weights are Apache-2.0, Copyright DiffuSpace and the Qwen team.
|