Update README.md
Browse files
README.md
CHANGED
|
@@ -21,7 +21,7 @@ from transformers import AutoTokenizer, AutoModelForCausalLM
|
|
| 21 |
|
| 22 |
tokenizer = AutoTokenizer.from_pretrained("budecosystem/Boomer-4b", trust_remote_code=True)
|
| 23 |
model = AutoModelForCausalLM.from_pretrained("budecosystem/Boomer-4b", torch_dtype=torch.bfloat16)
|
| 24 |
-
inputs = tokenizer("Quantum
|
| 25 |
sample = model.generate(**inputs, max_length=128)
|
| 26 |
print(tokenizer.decode(sample[0]))
|
| 27 |
```
|
|
@@ -56,7 +56,7 @@ Boomer-4b has been rigorously evaluated across several benchmarks:
|
|
| 56 |
|
| 57 |
| Model | MMLU | ARC | HellaSwag | GSM8K | Winogrande | MATH | MathQA | DROP | LogiQA |
|
| 58 |
|-------|------|-----|-----------|-------|------------|------|--------|------|--------|
|
| 59 |
-
| Boomer-4b | 55.59 | 58.53 | 74.70 | 47.76 | 72.22 | 4.00 | 35.98 | 0.74 | 31.80 |
|
| 60 |
| GeneZC/MiniChat-3B | 39.17 | 44.03 | 67.19 | 10.54 | 65.27 | - | - | - | - |
|
| 61 |
| openlm-research/open_11ama_3b_v2 | 27.12 | 44.03 | 71.6 | 0.91 | 67.01 | - | - | - | - |
|
| 62 |
| microsoft/phi-2 | 58.11 | 61.09 | 75.11 | 54.81 | 74.35 | - | - | - | - |
|
|
|
|
| 21 |
|
| 22 |
tokenizer = AutoTokenizer.from_pretrained("budecosystem/Boomer-4b", trust_remote_code=True)
|
| 23 |
model = AutoModelForCausalLM.from_pretrained("budecosystem/Boomer-4b", torch_dtype=torch.bfloat16)
|
| 24 |
+
inputs = tokenizer("Quantum mechanics is ", return_tensors="pt")
|
| 25 |
sample = model.generate(**inputs, max_length=128)
|
| 26 |
print(tokenizer.decode(sample[0]))
|
| 27 |
```
|
|
|
|
| 56 |
|
| 57 |
| Model | MMLU | ARC | HellaSwag | GSM8K | Winogrande | MATH | MathQA | DROP | LogiQA |
|
| 58 |
|-------|------|-----|-----------|-------|------------|------|--------|------|--------|
|
| 59 |
+
| **Boomer-4b** | 55.59 | 58.53 | **74.70** | 47.76 | **72.22** | 4.00 | 35.98 | 0.74 | 31.80 |
|
| 60 |
| GeneZC/MiniChat-3B | 39.17 | 44.03 | 67.19 | 10.54 | 65.27 | - | - | - | - |
|
| 61 |
| openlm-research/open_11ama_3b_v2 | 27.12 | 44.03 | 71.6 | 0.91 | 67.01 | - | - | - | - |
|
| 62 |
| microsoft/phi-2 | 58.11 | 61.09 | 75.11 | 54.81 | 74.35 | - | - | - | - |
|