feat: impl session snapshot, add comments for llama-rs parameters (#33)

Author: hlhr202

Cargo.lock

@@ -1,6 +1,6 @@
 {
   "name": "@llama-node/examples",
-  "version": "0.0.32",
+  "version": "0.0.33",
   "description": "",
   "main": "index.js",
   "type": "module",
@@ -18,9 +18,9 @@
     "langchain": "^0.0.56"
   },
   "dependencies": {
-    "@llama-node/core": "0.0.32",
-    "@llama-node/llama-cpp": "0.0.32",
-    "@llama-node/rwkv-cpp": "0.0.32",
-    "llama-node": "0.0.32"
+    "@llama-node/core": "0.0.33",
+    "@llama-node/llama-cpp": "0.0.33",
+    "@llama-node/rwkv-cpp": "0.0.33",
+    "llama-node": "0.0.33"
   }
 }

example/package.json

@@ -1,6 +1,6 @@
 {
   "name": "llama-node",
-  "version": "0.0.32",
+  "version": "0.0.33",
   "description": "Node.js Library for Large Language Model LLaMA/RWKV",
   "keywords": [
     "llama",
@@ -59,10 +59,10 @@
   "packageManager": "[email protected]",
   "homepage": "https://llama-node.vercel.app/",
   "devDependencies": {
-    "@llama-node/cli": "0.0.32",
-    "@llama-node/core": "0.0.32",
-    "@llama-node/llama-cpp": "0.0.32",
-    "@llama-node/rwkv-cpp": "0.0.32",
+    "@llama-node/cli": "0.0.33",
+    "@llama-node/core": "0.0.33",
+    "@llama-node/llama-cpp": "0.0.33",
+    "@llama-node/rwkv-cpp": "0.0.33",
     "@napi-rs/cli": "^2.15.2",
     "@tensorflow/tfjs-node": "^4.2.0",
     "@types/node": "^18.15.5",
@@ -80,17 +80,17 @@
     "vitest": "^0.30.1"
   },
   "dependencies": {
-    "@llama-node/cli": "0.0.32"
+    "@llama-node/cli": "0.0.33"
   },
   "optionalDependencies": {
-    "@llama-node/core": "0.0.32",
-    "@llama-node/llama-cpp": "0.0.32",
-    "@llama-node/rwkv-cpp": "0.0.32"
+    "@llama-node/core": "0.0.33",
+    "@llama-node/llama-cpp": "0.0.33",
+    "@llama-node/rwkv-cpp": "0.0.33"
   },
   "peerDependencies": {
-    "@llama-node/cli": "0.0.32",
-    "@llama-node/core": "0.0.32",
-    "@llama-node/llama-cpp": "0.0.32",
-    "@llama-node/rwkv-cpp": "0.0.32"
+    "@llama-node/cli": "0.0.33",
+    "@llama-node/core": "0.0.33",
+    "@llama-node/llama-cpp": "0.0.33",
+    "@llama-node/rwkv-cpp": "0.0.33"
   }
 }

package.json

@@ -1,6 +1,6 @@
 {
   "name": "@llama-node/cli",
-  "version": "0.0.32",
+  "version": "0.0.33",
   "description": "",
   "main": "index.js",
   "type": "commonjs",
@@ -12,7 +12,7 @@
     "test": "tsx src/index.ts"
   },
   "dependencies": {
-    "@llama-node/core": "0.0.32",
+    "@llama-node/core": "0.0.33",
     "yargs": "^17.7.1"
   },
   "author": "",

packages/cli/package.json

@@ -19,6 +19,9 @@ napi = { version = "2.12.4", default-features = false, features = ["napi6", "asy
 napi-derive = "2.12.3"
 tokio = { version = "1.26.0", features = ["full"] }
 futures = "0.3"
+zstd = {version = "0.12", default-features = false}
+anyhow = "1.0.70"
+bincode = "1.3.3"
 
 
 [build-dependencies]

packages/core/Cargo.toml

@@ -2,6 +2,7 @@ import { InferenceResultType, LLama } from "../index";
 import path from "path";
 
 const model = path.resolve(process.cwd(), "../../ggml-alpaca-7b-q4.bin");
+// const persistSession = path.resolve(process.cwd(), "./tmp/session.bin");
 
 LLama.enableLogger();
 
@@ -31,6 +32,7 @@ llama.inference(
         repeatLastN: 64,
         seed: 0,
         feedPrompt: true,
+        // persistSession,
     },
     (response) => {
         switch (response.type) {