newsletter commited on
Commit
caa5669
1 Parent(s): 52adb7a

Upload README.md with huggingface_hub

Browse files
Files changed (1) hide show
  1. README.md +25 -11
README.md CHANGED
@@ -13,7 +13,7 @@ tags:
13
  - synthetic data
14
  - llama-cpp
15
  - gguf-my-repo
16
- base_model: LeoLM/leo-mistral-hessianai-7b
17
  model-index:
18
  - name: DiscoLM_German_7b_v1
19
  results: []
@@ -22,29 +22,43 @@ model-index:
22
  # newsletter/DiscoLM_German_7b_v1-Q6_K-GGUF
23
  This model was converted to GGUF format from [`DiscoResearch/DiscoLM_German_7b_v1`](https://huggingface.co/DiscoResearch/DiscoLM_German_7b_v1) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
24
  Refer to the [original model card](https://huggingface.co/DiscoResearch/DiscoLM_German_7b_v1) for more details on the model.
25
- ## Use with llama.cpp
26
 
27
- Install llama.cpp through brew.
 
28
 
29
  ```bash
30
- brew install ggerganov/ggerganov/llama.cpp
 
31
  ```
32
  Invoke the llama.cpp server or the CLI.
33
 
34
- CLI:
35
-
36
  ```bash
37
- llama-cli --hf-repo newsletter/DiscoLM_German_7b_v1-Q6_K-GGUF --model discolm_german_7b_v1.Q6_K.gguf -p "The meaning to life and the universe is"
38
  ```
39
 
40
- Server:
41
-
42
  ```bash
43
- llama-server --hf-repo newsletter/DiscoLM_German_7b_v1-Q6_K-GGUF --model discolm_german_7b_v1.Q6_K.gguf -c 2048
44
  ```
45
 
46
  Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well.
47
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
48
  ```
49
- git clone https://github.com/ggerganov/llama.cpp && cd llama.cpp && make && ./main -m discolm_german_7b_v1.Q6_K.gguf -n 128
50
  ```
 
13
  - synthetic data
14
  - llama-cpp
15
  - gguf-my-repo
16
+ base_model: DiscoResearch/DiscoLM_German_7b_v1
17
  model-index:
18
  - name: DiscoLM_German_7b_v1
19
  results: []
 
22
  # newsletter/DiscoLM_German_7b_v1-Q6_K-GGUF
23
  This model was converted to GGUF format from [`DiscoResearch/DiscoLM_German_7b_v1`](https://huggingface.co/DiscoResearch/DiscoLM_German_7b_v1) using llama.cpp via the ggml.ai's [GGUF-my-repo](https://huggingface.co/spaces/ggml-org/gguf-my-repo) space.
24
  Refer to the [original model card](https://huggingface.co/DiscoResearch/DiscoLM_German_7b_v1) for more details on the model.
 
25
 
26
+ ## Use with llama.cpp
27
+ Install llama.cpp through brew (works on Mac and Linux)
28
 
29
  ```bash
30
+ brew install llama.cpp
31
+
32
  ```
33
  Invoke the llama.cpp server or the CLI.
34
 
35
+ ### CLI:
 
36
  ```bash
37
+ llama --hf-repo newsletter/DiscoLM_German_7b_v1-Q6_K-GGUF --hf-file discolm_german_7b_v1-q6_k.gguf -p "The meaning to life and the universe is"
38
  ```
39
 
40
+ ### Server:
 
41
  ```bash
42
+ llama-server --hf-repo newsletter/DiscoLM_German_7b_v1-Q6_K-GGUF --hf-file discolm_german_7b_v1-q6_k.gguf -c 2048
43
  ```
44
 
45
  Note: You can also use this checkpoint directly through the [usage steps](https://github.com/ggerganov/llama.cpp?tab=readme-ov-file#usage) listed in the Llama.cpp repo as well.
46
 
47
+ Step 1: Clone llama.cpp from GitHub.
48
+ ```
49
+ git clone https://github.com/ggerganov/llama.cpp
50
+ ```
51
+
52
+ Step 2: Move into the llama.cpp folder and build it with `LLAMA_CURL=1` flag along with other hardware-specific flags (for ex: LLAMA_CUDA=1 for Nvidia GPUs on Linux).
53
+ ```
54
+ cd llama.cpp && LLAMA_CURL=1 make
55
+ ```
56
+
57
+ Step 3: Run inference through the main binary.
58
+ ```
59
+ ./main --hf-repo newsletter/DiscoLM_German_7b_v1-Q6_K-GGUF --hf-file discolm_german_7b_v1-q6_k.gguf -p "The meaning to life and the universe is"
60
+ ```
61
+ or
62
  ```
63
+ ./server --hf-repo newsletter/DiscoLM_German_7b_v1-Q6_K-GGUF --hf-file discolm_german_7b_v1-q6_k.gguf -c 2048
64
  ```