vrr/prima.cpp
2
0
Fork 0
mirror of https://github.com/Lizonghang/prima.cpp.git synced 2025-09-07 14:19:03 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

update README

Browse source
This commit is contained in:
Li, Zonghang 2025-06-27 20:16:30 +04:00
parent 11ce0d58f7
commit e8d3e5a631
2 changed files with 6 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

8
README.md
View file

@ -232,10 +232,10 @@ Assume we have a host machine with at least 32 CPU cores, 32 GiB RAM, and 32 GiB
1. Pull our prebuilt Docker image (e.g., [`prima.cpp:1.0.2-cuda`](https://hub.docker.com/repository/docker/lizonghango00o1/prima.cpp/general)) and run 4 containers: 1. Pull our prebuilt Docker image (e.g., [`prima.cpp:1.0.2-cuda`](https://hub.docker.com/repository/docker/lizonghango00o1/prima.cpp/general)) and run 4 containers:
```shell ```shell
sudo docker run -dit --name prima-v1 --memory=8gb --memory-swap=8gb --cpus 8 --cpuset-cpus="0-7" --network host --gpus all prima.cpp:1.0.1-cuda sudo docker run -dit --name prima-v1 --memory=8gb --memory-swap=8gb --cpus 8 --cpuset-cpus="0-7" --network host --gpus all prima.cpp:1.0.2-cuda
sudo docker run -dit --name prima-v2 --memory=8gb --memory-swap=8gb --cpus 8 --cpuset-cpus="8-15" --network host --gpus all prima.cpp:1.0.1-cuda sudo docker run -dit --name prima-v2 --memory=8gb --memory-swap=8gb --cpus 8 --cpuset-cpus="8-15" --network host --gpus all prima.cpp:1.0.2-cuda
sudo docker run -dit --name prima-v3 --memory=8gb --memory-swap=8gb --cpus 8 --cpuset-cpus="16-23" --network host --gpus all prima.cpp:1.0.1-cuda sudo docker run -dit --name prima-v3 --memory=8gb --memory-swap=8gb --cpus 8 --cpuset-cpus="16-23" --network host --gpus all prima.cpp:1.0.2-cuda
sudo docker run -dit --name prima-v4 --memory=8gb --memory-swap=8gb --cpus 8 --cpuset-cpus="24-31" --network host --gpus all prima.cpp:1.0.1-cuda sudo docker run -dit --name prima-v4 --memory=8gb --memory-swap=8gb --cpus 8 --cpuset-cpus="24-31" --network host --gpus all prima.cpp:1.0.2-cuda
``` ```
2. Download the model file [`qwq-32b-q4_k_m.gguf`](https://huggingface.co/Qwen/QwQ-32B-GGUF) and copy it into each container: 2. Download the model file [`qwq-32b-q4_k_m.gguf`](https://huggingface.co/Qwen/QwQ-32B-GGUF) and copy it into each container:

4
src/llama.cpp
View file

@ -21365,8 +21365,6 @@ void * llama_context_setup_backend(
|| model->n_gpu_layers == 0) { || model->n_gpu_layers == 0) {
continue; continue;
} }
#elif defined(GGML_USE_METAL)
#endif #endif
ok = ok & ggml_backend_sched_reserve(ctx->sched[i], gf[i]); ok = ok & ggml_backend_sched_reserve(ctx->sched[i], gf[i]);
@ -21933,9 +21931,9 @@ void llama_model_compute_buf_size(
const int64_t n_result = hparams.n_vocab * cparams.n_ubatch; const int64_t n_result = hparams.n_vocab * cparams.n_ubatch;
// weights // weights
const int64_t nb_output_w = n_bytes.nb_output_w;
const int64_t nb_attn_norm_w = n_bytes.nb_attn_norm_w; const int64_t nb_attn_norm_w = n_bytes.nb_attn_norm_w;
const int64_t nb_attn_q_w = n_bytes.nb_attn_q_w; const int64_t nb_attn_q_w = n_bytes.nb_attn_q_w;
// const int64_t nb_output_w = n_bytes.nb_output_w;
// format bytes // format bytes
const int64_t type_size_f32 = ggml_type_size(GGML_TYPE_F32); const int64_t type_size_f32 = ggml_type_size(GGML_TYPE_F32);