@tuned-tensor/cli 0.26.2 → 0.28.0

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.
package/README.md CHANGED
@@ -47,6 +47,20 @@ hidden in the local-only release are available under `tt cloud`.
47
47
 
48
48
  ## Install
49
49
 
50
+ On Windows (PowerShell 5.1 or 7), with Node.js 22.19+ installed:
51
+
52
+ ```powershell
53
+ winget install --id OpenJS.NodeJS.LTS --exact --source winget # if Node is missing
54
+ # Open a new terminal after installing Node.
55
+ npm.cmd install -g --ignore-scripts @tuned-tensor/cli
56
+ tt.cmd --version
57
+ ```
58
+
59
+ The [Windows guide](docs/windows.md) covers native setup with a bundled Node
60
+ runtime, PowerShell execution policies, upgrades, and the WinGet distribution
61
+ process. `TunedTensor.CLI` will become installable through WinGet after the first
62
+ Windows release and catalog submission are accepted.
63
+
50
64
  On Linux or macOS:
51
65
 
52
66
  ```bash
@@ -303,8 +317,12 @@ models or certify their fit. Old snapshots without a decision probe show
303
317
  | Native Windows x64 | CPU path; end-to-end training not yet certified | Not certified natively; check a Linux/WSL2 environment or use AWS | Linux/CUDA host |
304
318
  | Intel macOS / Windows ARM64 | No compatible wheels in the current lockfiles | Remote AWS GPU | Remote Linux/CUDA host |
305
319
 
306
- The native Windows CI job verifies hardware discovery and displayed guidance,
307
- not model training. WSL is assessed as its own Linux environment. AMD/Intel GPU
320
+ The native Windows CI job verifies installation, PowerShell CLI commands,
321
+ pipeline previews, hardware discovery and displayed guidance, not model training.
322
+ Agent-approved spec creation, editing and history work on native Windows and Linux.
323
+ AWS transport/background training should run inside WSL2/Linux. See
324
+ [Windows limitations](docs/windows.md#native-cli-scope).
325
+ WSL is assessed as its own Linux environment. AMD/Intel GPU
308
326
  acceleration is not enabled by the packaged CUDA runtime. Decision resource
309
327
  warnings use conservative advisory thresholds (8 GiB available working memory,
310
328
  6 GiB free model-cache disk, 4 CPU cores); they are not measured model minima.
@@ -861,3 +879,11 @@ the remaining steps toward the full conversational product promise.
861
879
  ## License
862
880
 
863
881
  Apache-2.0
882
+
883
+ ## Serverless decision inference
884
+
885
+ `tt cloud endpoints` uploads complete local Laya decision checkpoints to an
886
+ authenticated CPU endpoint, with per-request prepaid billing. See the
887
+ [hosting guide](docs/local-runtime/hosting.md) for creation, inference keys,
888
+ retry-safe requests, costs and deletion. Hosting requires the new server API;
889
+ GPU and adapter checkpoints remain outside this serverless flow.