f16 changed to f32
Browse files
MODELFILE
CHANGED
|
@@ -1,25 +1,25 @@
|
|
| 1 |
-
# MODELFILE for Qwen3-Coder-30B-A3B-Instruct-GGUF
|
| 2 |
-
# Used by LM Studio, OpenWebUI, GPT4All, etc.
|
| 3 |
-
|
| 4 |
-
context_length: 32768
|
| 5 |
-
embedding: false
|
| 6 |
-
|
| 7 |
-
|
| 8 |
-
# Chat template using ChatML (used by Qwen)
|
| 9 |
-
prompt_template: >-
|
| 10 |
-
<|im_start|>system
|
| 11 |
-
You are a helpful assistant.<|im_end|>
|
| 12 |
-
<|im_start|>user
|
| 13 |
-
{prompt}<|im_end|>
|
| 14 |
-
<|im_start|>assistant
|
| 15 |
-
|
| 16 |
-
# Stop sequences help end generation cleanly
|
| 17 |
-
stop: "<|im_end|>"
|
| 18 |
-
stop: "<|im_start|>"
|
| 19 |
-
|
| 20 |
-
# Default sampling
|
| 21 |
-
temperature: 0.6
|
| 22 |
-
top_p: 0.95
|
| 23 |
-
top_k: 20
|
| 24 |
-
min_p: 0.0
|
| 25 |
-
repeat_penalty: 1.1
|
|
|
|
| 1 |
+
# MODELFILE for Qwen3-Coder-30B-A3B-Instruct-GGUF
|
| 2 |
+
# Used by LM Studio, OpenWebUI, GPT4All, etc.
|
| 3 |
+
|
| 4 |
+
context_length: 32768
|
| 5 |
+
embedding: false
|
| 6 |
+
f32: cpu
|
| 7 |
+
|
| 8 |
+
# Chat template using ChatML (used by Qwen)
|
| 9 |
+
prompt_template: >-
|
| 10 |
+
<|im_start|>system
|
| 11 |
+
You are a helpful assistant.<|im_end|>
|
| 12 |
+
<|im_start|>user
|
| 13 |
+
{prompt}<|im_end|>
|
| 14 |
+
<|im_start|>assistant
|
| 15 |
+
|
| 16 |
+
# Stop sequences help end generation cleanly
|
| 17 |
+
stop: "<|im_end|>"
|
| 18 |
+
stop: "<|im_start|>"
|
| 19 |
+
|
| 20 |
+
# Default sampling
|
| 21 |
+
temperature: 0.6
|
| 22 |
+
top_p: 0.95
|
| 23 |
+
top_k: 20
|
| 24 |
+
min_p: 0.0
|
| 25 |
+
repeat_penalty: 1.1
|