@tuned-tensor/cli 0.13.1-beta.7 → 0.13.1

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
@@ -43,9 +43,9 @@ resource defaults, data contract, and real CUDA acceptance test.
43
43
  TT Local currently ships two such methods:
44
44
 
45
45
  - **Adapter** (`training/adapter`): certified Transformers checkpoints and
46
- PEFT LoRA SFT. Driven by `tt run` and adapter pipelines.
46
+ PEFT LoRA SFT. Driven by `tt pipeline run`.
47
47
  - **Foundation** (`training/foundation`): a from-scratch BPE tokenizer and GPT
48
- with no Transformers/PEFT. Driven only by `tt pipeline` from a foundation
48
+ with no Transformers/PEFT. Driven by `tt pipeline run` from a foundation
49
49
  `tunedtensor.json`. Pretrain, SFT, and RL require CUDA. Tokenizer training
50
50
  and the CPU unit tests do not. The current runtime is single-process and
51
51
  rejects `nproc_per_node` values above 1 and unsupported steps such as
@@ -108,7 +108,7 @@ tt doctor tunedtensor.json
108
108
  tt validate tunedtensor.json
109
109
  tt models prefetch tunedtensor.json
110
110
  tt models verify-base tunedtensor.json
111
- tt run tunedtensor.json
111
+ tt pipeline run --spec tunedtensor.json --config local-runner.json
112
112
  ```
113
113
 
114
114
  `doctor` resolves the same bundled project and paths the run will use, imports
package/package.json CHANGED
@@ -1,6 +1,6 @@
1
1
  {
2
2
  "name": "@tuned-tensor/cli",
3
- "version": "0.13.1-beta.7",
3
+ "version": "0.13.1",
4
4
  "description": "Local TT agent and CUDA fine-tuning CLI for Tuned Tensor workflows",
5
5
  "repository": {
6
6
  "type": "git",