Commit
•
35b1d48
1
Parent(s):
7f137f4
Upload README.md
Browse files
README.md
CHANGED
|
@@ -7,7 +7,7 @@ tags:
|
|
| 7 |
- moe
|
| 8 |
---
|
| 9 |
|
| 10 |
-
### Mini-Jamba
|
| 11 |
|
| 12 |
[**Experimental Version**] We initialized the model according to [Jamba](https://huggingface.co/ai21labs/Jamba-v0.1), but with much smaller parameters. It was then trained using about 1B of python code, and has the simplest python code generation capabilities.
|
| 13 |
|
|
@@ -31,12 +31,12 @@ prompt = '''def min(arr):
|
|
| 31 |
'''
|
| 32 |
|
| 33 |
tokenizer = AutoTokenizer.from_pretrained(
|
| 34 |
-
"TechxGenus/Mini-Jamba",
|
| 35 |
trust_remote_code=True,
|
| 36 |
)
|
| 37 |
tokenizer.pad_token = tokenizer.eos_token
|
| 38 |
model = AutoModelForCausalLM.from_pretrained(
|
| 39 |
-
"TechxGenus/Mini-Jamba",
|
| 40 |
torch_dtype=torch.float16,
|
| 41 |
device_map="auto",
|
| 42 |
trust_remote_code=True,
|
|
|
|
| 7 |
- moe
|
| 8 |
---
|
| 9 |
|
| 10 |
+
### Mini-Jamba-v2
|
| 11 |
|
| 12 |
[**Experimental Version**] We initialized the model according to [Jamba](https://huggingface.co/ai21labs/Jamba-v0.1), but with much smaller parameters. It was then trained using about 1B of python code, and has the simplest python code generation capabilities.
|
| 13 |
|
|
|
|
| 31 |
'''
|
| 32 |
|
| 33 |
tokenizer = AutoTokenizer.from_pretrained(
|
| 34 |
+
"TechxGenus/Mini-Jamba-v2",
|
| 35 |
trust_remote_code=True,
|
| 36 |
)
|
| 37 |
tokenizer.pad_token = tokenizer.eos_token
|
| 38 |
model = AutoModelForCausalLM.from_pretrained(
|
| 39 |
+
"TechxGenus/Mini-Jamba-v2",
|
| 40 |
torch_dtype=torch.float16,
|
| 41 |
device_map="auto",
|
| 42 |
trust_remote_code=True,
|