morriszms commited on
Commit
532fca9
1 Parent(s): f843c69

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +20 -12
README.md CHANGED
@@ -25,8 +25,16 @@ This repo contains GGUF format model files for [deepseek-ai/deepseek-coder-1.3b-
25
 
26
  The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4011](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
27
 
 
 
 
 
 
 
 
28
  ## Prompt template
29
 
 
30
  ```
31
  <|begin▁of▁sentence|>{system_prompt}### Instruction:
32
  {prompt}
@@ -37,18 +45,18 @@ The files were quantized using machines provided by [TensorBlock](https://tensor
37
 
38
  | Filename | Quant type | File Size | Description |
39
  | -------- | ---------- | --------- | ----------- |
40
- | [deepseek-coder-1.3b-instruct-Q2_K.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q2_K.gguf) | Q2_K | 0.521 GB | smallest, significant quality loss - not recommended for most purposes |
41
- | [deepseek-coder-1.3b-instruct-Q3_K_S.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q3_K_S.gguf) | Q3_K_S | 0.598 GB | very small, high quality loss |
42
- | [deepseek-coder-1.3b-instruct-Q3_K_M.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q3_K_M.gguf) | Q3_K_M | 0.656 GB | very small, high quality loss |
43
- | [deepseek-coder-1.3b-instruct-Q3_K_L.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q3_K_L.gguf) | Q3_K_L | 0.693 GB | small, substantial quality loss |
44
- | [deepseek-coder-1.3b-instruct-Q4_0.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q4_0.gguf) | Q4_0 | 0.723 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
45
- | [deepseek-coder-1.3b-instruct-Q4_K_S.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q4_K_S.gguf) | Q4_K_S | 0.758 GB | small, greater quality loss |
46
- | [deepseek-coder-1.3b-instruct-Q4_K_M.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q4_K_M.gguf) | Q4_K_M | 0.813 GB | medium, balanced quality - recommended |
47
- | [deepseek-coder-1.3b-instruct-Q5_0.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q5_0.gguf) | Q5_0 | 0.872 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
48
- | [deepseek-coder-1.3b-instruct-Q5_K_S.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q5_K_S.gguf) | Q5_K_S | 0.887 GB | large, low quality loss - recommended |
49
- | [deepseek-coder-1.3b-instruct-Q5_K_M.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q5_K_M.gguf) | Q5_K_M | 0.933 GB | large, very low quality loss - recommended |
50
- | [deepseek-coder-1.3b-instruct-Q6_K.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q6_K.gguf) | Q6_K | 1.091 GB | very large, extremely low quality loss |
51
- | [deepseek-coder-1.3b-instruct-Q8_0.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/tree/main/deepseek-coder-1.3b-instruct-Q8_0.gguf) | Q8_0 | 1.334 GB | very large, extremely low quality loss - not recommended |
52
 
53
 
54
  ## Downloading instruction
 
25
 
26
  The files were quantized using machines provided by [TensorBlock](https://tensorblock.co/), and they are compatible with llama.cpp as of [commit b4011](https://github.com/ggerganov/llama.cpp/commit/a6744e43e80f4be6398fc7733a01642c846dce1d).
27
 
28
+
29
+ <div style="text-align: left; margin: 20px 0;">
30
+ <a href="https://tensorblock.co/waitlist/client" style="display: inline-block; padding: 10px 20px; background-color: #007bff; color: white; text-decoration: none; border-radius: 5px; font-weight: bold;">
31
+ Run them on the TensorBlock client using your local machine ↗
32
+ </a>
33
+ </div>
34
+
35
  ## Prompt template
36
 
37
+
38
  ```
39
  <|begin▁of▁sentence|>{system_prompt}### Instruction:
40
  {prompt}
 
45
 
46
  | Filename | Quant type | File Size | Description |
47
  | -------- | ---------- | --------- | ----------- |
48
+ | [deepseek-coder-1.3b-instruct-Q2_K.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q2_K.gguf) | Q2_K | 0.521 GB | smallest, significant quality loss - not recommended for most purposes |
49
+ | [deepseek-coder-1.3b-instruct-Q3_K_S.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q3_K_S.gguf) | Q3_K_S | 0.598 GB | very small, high quality loss |
50
+ | [deepseek-coder-1.3b-instruct-Q3_K_M.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q3_K_M.gguf) | Q3_K_M | 0.656 GB | very small, high quality loss |
51
+ | [deepseek-coder-1.3b-instruct-Q3_K_L.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q3_K_L.gguf) | Q3_K_L | 0.693 GB | small, substantial quality loss |
52
+ | [deepseek-coder-1.3b-instruct-Q4_0.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q4_0.gguf) | Q4_0 | 0.723 GB | legacy; small, very high quality loss - prefer using Q3_K_M |
53
+ | [deepseek-coder-1.3b-instruct-Q4_K_S.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q4_K_S.gguf) | Q4_K_S | 0.758 GB | small, greater quality loss |
54
+ | [deepseek-coder-1.3b-instruct-Q4_K_M.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q4_K_M.gguf) | Q4_K_M | 0.813 GB | medium, balanced quality - recommended |
55
+ | [deepseek-coder-1.3b-instruct-Q5_0.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q5_0.gguf) | Q5_0 | 0.872 GB | legacy; medium, balanced quality - prefer using Q4_K_M |
56
+ | [deepseek-coder-1.3b-instruct-Q5_K_S.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q5_K_S.gguf) | Q5_K_S | 0.887 GB | large, low quality loss - recommended |
57
+ | [deepseek-coder-1.3b-instruct-Q5_K_M.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q5_K_M.gguf) | Q5_K_M | 0.933 GB | large, very low quality loss - recommended |
58
+ | [deepseek-coder-1.3b-instruct-Q6_K.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q6_K.gguf) | Q6_K | 1.091 GB | very large, extremely low quality loss |
59
+ | [deepseek-coder-1.3b-instruct-Q8_0.gguf](https://huggingface.co/tensorblock/deepseek-coder-1.3b-instruct-GGUF/blob/main/deepseek-coder-1.3b-instruct-Q8_0.gguf) | Q8_0 | 1.334 GB | very large, extremely low quality loss - not recommended |
60
 
61
 
62
  ## Downloading instruction