Upload LoRA adapter trained with Tinker

Base model: meta-llama/Llama-3.1-8B
Tinker checkpoint: tinker://7f4705e7-551f-5133-b4bb-33444c0c405b:train:0/sampler_weights/test-push-to-hub
Uploaded: 2025-12-28T22:36:55.297557

README.md

@@ -5,6 +5,7 @@ tags:
 - tinker
 - lora
 - sl
+- math
 license: llama3.1
 ---
 
@@ -39,16 +40,22 @@ tokenizer = AutoTokenizer.from_pretrained("meta-llama/Llama-3.1-8B")
 model = PeftModel.from_pretrained(base_model, "joschu0/tinker-llama-lora-test")
 ```
 
-### With Tinker (for continued training)
+### With Tinker (for sampling/inference)
+
+This checkpoint is also available on Tinker for high-throughput sampling:
 
 ```python
 import tinker
 
 sc = tinker.ServiceClient()
-# Note: The checkpoint must be published on Tinker for this to work
-training_client = sc.create_training_client_from_state("tinker://7f4705e7-551f-5133-b4bb-33444c0c405b:train:0/sampler_weights/test-push-to-hub")
+sampling_client = sc.create_sampling_client("tinker://7f4705e7-551f-5133-b4bb-33444c0c405b:train:0/sampler_weights/test-push-to-hub")
+result = sampling_client.sample(...)
 ```
 
+**Tinker path:** `tinker://7f4705e7-551f-5133-b4bb-33444c0c405b:train:0/sampler_weights/test-push-to-hub`
+
+> **Note:** This is a sampler checkpoint and can only be used for inference, not for continued training.
+
 ## Training
 
 This model was trained using the Tinker API. For more information about training