Text Generation
Transformers
Safetensors
qwen3
Generated from Trainer
grpo
trl
conversational
text-generation-inference
Instructions to use cs-552-2026-clankers-builder/math_model with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Transformers
How to use cs-552-2026-clankers-builder/math_model with Transformers:
# Use a pipeline as a high-level helper from transformers import pipeline pipe = pipeline("text-generation", model="cs-552-2026-clankers-builder/math_model") messages = [ {"role": "user", "content": "Who are you?"}, ] pipe(messages)# Load model directly from transformers import AutoTokenizer, AutoModelForCausalLM tokenizer = AutoTokenizer.from_pretrained("cs-552-2026-clankers-builder/math_model") model = AutoModelForCausalLM.from_pretrained("cs-552-2026-clankers-builder/math_model", device_map="auto") messages = [ {"role": "user", "content": "Who are you?"}, ] inputs = tokenizer.apply_chat_template( messages, add_generation_prompt=True, tokenize=True, return_dict=True, return_tensors="pt", ).to(model.device) outputs = model.generate(**inputs, max_new_tokens=40) print(tokenizer.decode(outputs[0][inputs["input_ids"].shape[-1]:])) - Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- vLLM
How to use cs-552-2026-clankers-builder/math_model with vLLM:
Install from pip and serve model
# Install vLLM from pip: pip install vllm # Start the vLLM server: vllm serve "cs-552-2026-clankers-builder/math_model" # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:8000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "cs-552-2026-clankers-builder/math_model", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker
docker model run hf.co/cs-552-2026-clankers-builder/math_model
- SGLang
How to use cs-552-2026-clankers-builder/math_model with SGLang:
Install from pip and serve model
# Install SGLang from pip: pip install sglang # Start the SGLang server: python3 -m sglang.launch_server \ --model-path "cs-552-2026-clankers-builder/math_model" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "cs-552-2026-clankers-builder/math_model", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }'Use Docker images
docker run --gpus all \ --shm-size 32g \ -p 30000:30000 \ -v ~/.cache/huggingface:/root/.cache/huggingface \ --env "HF_TOKEN=<secret>" \ --ipc=host \ lmsysorg/sglang:latest \ python3 -m sglang.launch_server \ --model-path "cs-552-2026-clankers-builder/math_model" \ --host 0.0.0.0 \ --port 30000 # Call the server using curl (OpenAI-compatible API): curl -X POST "http://localhost:30000/v1/chat/completions" \ -H "Content-Type: application/json" \ --data '{ "model": "cs-552-2026-clankers-builder/math_model", "messages": [ { "role": "user", "content": "What is the capital of France?" } ] }' - Docker Model Runner
How to use cs-552-2026-clankers-builder/math_model with Docker Model Runner:
docker model run hf.co/cs-552-2026-clankers-builder/math_model
Update Automated MNLP evaluation report (2026-05-21)
Browse files- EVAL_REPORT.md +17 -36
EVAL_REPORT.md
CHANGED
|
@@ -2,7 +2,7 @@
|
|
| 2 |
|
| 3 |
- **Model repo:** [`cs-552-2026-clankers-builder/math_model`](https://huggingface.co/cs-552-2026-clankers-builder/math_model)
|
| 4 |
- **Owner(s):** group **clankers-builder**
|
| 5 |
-
- **Generated at:** 2026-05-
|
| 6 |
- **Pipeline:** [mnlp-project-ci](https://github.com/eric11eca/mnlp-project-ci)
|
| 7 |
|
| 8 |
_This PR is opened automatically by the course CI. It is **non-blocking** — you do not need to merge it. The next nightly run will refresh this file._
|
|
@@ -24,47 +24,34 @@ _Prompts are intentionally omitted to avoid revealing benchmark contents. For mu
|
|
| 24 |
|
| 25 |
**Correct** (1 shown)
|
| 26 |
|
| 27 |
-
- **reference**: `
|
| 28 |
- **overall** (1/8 completions correct)
|
| 29 |
-
- **extracted** (✓): `
|
| 30 |
- **completion**:
|
| 31 |
|
| 32 |
```text
|
| 33 |
<think>
|
| 34 |
-
Okay,
|
| 35 |
|
| 36 |
-
First, let me
|
| 37 |
|
| 38 |
... [truncated 823 chars] ...
|
| 39 |
|
| 40 |
-
|
| 41 |
|
| 42 |
-
|
| 43 |
|
| 44 |
-
|
| 45 |
|
| 46 |
-
|
| 47 |
|
| 48 |
-
|
| 49 |
-
|
| 50 |
-
Calculating 3*53.25: 53*3=159, 0.25*3=0.75, so 159+0.75=159.75
|
| 51 |
-
|
| 52 |
-
Then 159.75=10x => x=159.75/10=15.975
|
| 53 |
-
|
| 54 |
-
Convert 15.975 to a fraction: 15.975=15 +0.975
|
| 55 |
-
|
| 56 |
-
0.975=975/1000=39/40
|
| 57 |
-
|
| 58 |
-
Therefore, 15.975=15 39/40=639/40
|
| 59 |
-
|
| 60 |
-
Yes, that all checks out. So, the answer is 639/40 inches.
|
| 61 |
|
| 62 |
**Final Answer**
|
| 63 |
-
|
| 64 |
-
|
| 65 |
</think>
|
| 66 |
|
| 67 |
-
\boxed{
|
| 68 |
```
|
| 69 |
|
| 70 |
**Incorrect** (1 shown)
|
|
@@ -76,19 +63,13 @@ _Prompts are intentionally omitted to avoid revealing benchmark contents. For mu
|
|
| 76 |
|
| 77 |
```text
|
| 78 |
<think>
|
| 79 |
-
Okay, so Derek starts with a debt of π MIT, which is approximately 3.1416 MIT. Each day he flips a fair coin. If it's heads, he reduces his debt by 1 MIT, and if it's tails, he reduces it by 2 MITs. If his debt ever becomes nonpositive, he's debt-free. Otherwise, after the payment, his debt
|
| 80 |
-
|
| 81 |
-
... [truncated 823 chars] ...
|
| 82 |
-
|
| 83 |
-
: 2^1 -1 =1. P(1)=1/2. Hmm, not quite.
|
| 84 |
-
|
| 85 |
-
Wait, maybe the numerator is (2^{n} - 1)/something. Let's see:
|
| 86 |
|
| 87 |
-
|
| 88 |
|
| 89 |
-
|
| 90 |
|
| 91 |
-
|
| 92 |
|
| 93 |
-
|
| 94 |
```
|
|
|
|
| 2 |
|
| 3 |
- **Model repo:** [`cs-552-2026-clankers-builder/math_model`](https://huggingface.co/cs-552-2026-clankers-builder/math_model)
|
| 4 |
- **Owner(s):** group **clankers-builder**
|
| 5 |
+
- **Generated at:** 2026-05-21T03:28:21+00:00 (UTC)
|
| 6 |
- **Pipeline:** [mnlp-project-ci](https://github.com/eric11eca/mnlp-project-ci)
|
| 7 |
|
| 8 |
_This PR is opened automatically by the course CI. It is **non-blocking** — you do not need to merge it. The next nightly run will refresh this file._
|
|
|
|
| 24 |
|
| 25 |
**Correct** (1 shown)
|
| 26 |
|
| 27 |
+
- **reference**: `106`
|
| 28 |
- **overall** (1/8 completions correct)
|
| 29 |
+
- **extracted** (✓): `106`
|
| 30 |
- **completion**:
|
| 31 |
|
| 32 |
```text
|
| 33 |
<think>
|
| 34 |
+
Okay, let me try to figure out this problem. So, we have this product from k=4 to k=63 of [log base k of (5^{k² -1}) divided by log base (k+1) of (5^{k² -4})]. And we need to simplify this product and express it as m/n where m and n are coprime, then find m + n. Hmm, that seems complicated, but maybe there's a pattern or some telescoping product here.
|
| 35 |
|
| 36 |
+
First, let me recall that log_b(a) ca
|
| 37 |
|
| 38 |
... [truncated 823 chars] ...
|
| 39 |
|
| 40 |
+
)/ln(4) = 3 ln4 / ln4 = 3. Correct.
|
| 41 |
|
| 42 |
+
First product: Product_{k=4}^{63} [(k² -1)/(k² -4)] = [Product (k -1)/(k -2)] * [Product (k +1)/(k +2)] = [Product (k -1)/(k -2)] = 62/2 = 31. Wait, but 62/2 is 31, yes. And [Product (k +1)/(k +2)] = 5/65 = 1/13. So 31 * 1/13 = 31/13. Then multiply by 3: 31/13 * 3 = 93/13. Then 93 and 13: 13*7=91, 93-91=2, so 93/13 is reduced. So m=93, n=13, m + n=106.
|
| 43 |
|
| 44 |
+
But let me check the first product again. The product of (k -1)/(k -2) from k=4 to 63. Let's see: when k=4, (4-1)/(4-2)=3/2; k=5:4/3; k=6:5/4; ... up to k=63:62/61. So the product is (3/2)*(4/3)*(5/4)*...*(62/61). All the terms cancel except the first denominator and the last numerator: 62/2 = 31. Correct.
|
| 45 |
|
| 46 |
+
Similarly, the product of (k +1)/(k +2) from k=4 to 63. When k=4, (5/6); k=5:6/7; ... up to k=63:64/65. So the product is (5/6)*(6/7)*...*(64/65). All terms cancel except the first numerator and the last denominator: 5/65 = 1/13. Correct. So 31 * 1/13 = 31/13. Then multiplied by 3 gives 93/13. Correct.
|
| 47 |
|
| 48 |
+
So 93/13 is in lowest terms because 93 divided by 13 is 7 with a remainder of 2. So yes, m=93, n=13, m + n=106. Therefore, the answer is 106.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 49 |
|
| 50 |
**Final Answer**
|
| 51 |
+
\boxed{106}
|
|
|
|
| 52 |
</think>
|
| 53 |
|
| 54 |
+
\boxed{106}
|
| 55 |
```
|
| 56 |
|
| 57 |
**Incorrect** (1 shown)
|
|
|
|
| 63 |
|
| 64 |
```text
|
| 65 |
<think>
|
| 66 |
+
Okay, so Derek starts with a debt of π MIT, which is approximately 3.1416 MIT. Each day he flips a fair coin. If it's heads, he reduces his debt by 1 MIT, and if it's tails, he reduces it by 2 MITs. If his debt ever becomes nonpositive, he's debt-free. Otherwise, after the payment, his debt is doubled. We need to find the probability that Derek eventually becomes debt-free.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
|
| 68 |
+
Hmm, this see
|
| 69 |
|
| 70 |
+
... [truncated 822 chars] ...
|
| 71 |
|
| 72 |
+
Similarly, for x = 2, P(2) = 0.5*P(2*(2 - 1)) + 0.5*P(2*(2 - 2)) = 0.5*P(2) + 0.5*P(0). Therefore, P(2) = 0.5*P(2) + 0.5*1. Solving for P(2), we get 0.5*P(2) = 0.5*1 => P(2) = 1. Similarly, for x = 3, P(3) = 0.5*P(2*(3 - 1)) + 0.5*P(2*(3 - 2)) = 0.5*P(4) + 0.5*P(2). But P(4) = 1, P(2) = 1, so P(3) = 0.5*1 + 0.5*1 = 1. So again, P(x) = 1.
|
| 73 |
|
| 74 |
+
Therefore, it seems that P(x) = 1 for all x >= 0. Therefore, the probability that Derek eventually becomes debt-free is 1. But that seems surprising. Let me test with a smaller example. Suppose Derek starts with a debt of 1. Then, regardless of whether he pays 1 or 2, he becomes debt-free. So P(1) = 1. If he starts with a debt of 2, then he can pay 1, leading to 1, which is debt-free, or pay 2, leading to 0, which is debt-free. So again, P(2) = 1. If he starts with a debt of 3, he can pay 1, leading to 2, which is debt-free (since after paying 1, the debt is 2, which is positive, so it's doubled to 4. Then, he pays 1 again, leading to 3, which is debt-free (after paying 1, it's 3, which is positive, so doubled to 6. Then, he pays 1 again, leading to 5, doubled to 10. This seems like an infinite loop? Wait, but according to our previous reasoning
|
| 75 |
```
|