Release 2.16

.github/workflows/checks.yml

@@ -71,6 +71,7 @@ jobs:
           poetry check --lock
           poetry install --with dev
       - name: Authenticate HuggingFace CLI
+        if: env.HF_TOKEN != ''
         run: |
           pip install huggingface_hub
           huggingface-cli login --token "$HF_TOKEN"
@@ -116,6 +117,7 @@ jobs:
       - name: Type check
         run: poetry run mypy .
       - name: Authenticate HuggingFace CLI
+        if: env.HF_TOKEN != ''
         run: |
           pip install huggingface_hub
           huggingface-cli login --token "$HF_TOKEN"
@@ -212,6 +214,7 @@ jobs:
           name: test-coverage
           path: docs/source/_static/coverage
       - name: Authenticate HuggingFace CLI
+        if: env.HF_TOKEN != ''
         run: |
           pip install huggingface_hub
           huggingface-cli login --token "$HF_TOKEN"

demos/Colab_Compatibility.ipynb

@@ -2,7 +2,7 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": 3,
    "metadata": {},
    "outputs": [
     {
@@ -16,9 +16,9 @@
      "name": "stderr",
      "output_type": "stream",
      "text": [
-      "/var/folders/pr/77j77_bs2gl2stxyrvr14x3c0000gn/T/ipykernel_60991/3507779555.py:18: DeprecationWarning: `magic(...)` is deprecated since IPython 0.13 (warning added in 8.1), use run_line_magic(magic_name, parameter_s).\n",
+      "/var/folders/m3/z6c6rcdj1rbb2jh9vqpgvxg40000gn/T/ipykernel_86391/3507779555.py:18: DeprecationWarning: `magic(...)` is deprecated since IPython 0.13 (warning added in 8.1), use run_line_magic(magic_name, parameter_s).\n",
       "  ipython.magic(\"load_ext autoreload\")\n",
-      "/var/folders/pr/77j77_bs2gl2stxyrvr14x3c0000gn/T/ipykernel_60991/3507779555.py:19: DeprecationWarning: `magic(...)` is deprecated since IPython 0.13 (warning added in 8.1), use run_line_magic(magic_name, parameter_s).\n",
+      "/var/folders/m3/z6c6rcdj1rbb2jh9vqpgvxg40000gn/T/ipykernel_86391/3507779555.py:19: DeprecationWarning: `magic(...)` is deprecated since IPython 0.13 (warning added in 8.1), use run_line_magic(magic_name, parameter_s).\n",
       "  ipython.magic(\"autoreload 2\")\n"
      ]
     }
@@ -58,20 +58,19 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 2,
+   "execution_count": 4,
    "metadata": {},
    "outputs": [
     {
      "name": "stdout",
      "output_type": "stream",
      "text": [
-      "TransformerLens currently supports 211 models out of the box.\n"
+      "TransformerLens currently supports 216 models out of the box.\n"
      ]
     }
    ],
    "source": [
     "import torch\n",
-    "\n",
     "from transformer_lens import HookedTransformer, HookedEncoderDecoder, HookedEncoder, BertNextSentencePrediction, loading\n",
     "from transformers import AutoTokenizer, LlamaForCausalLM, LlamaTokenizer\n",
     "from typing import List\n",
@@ -89,7 +88,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 3,
+   "execution_count": 5,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -199,7 +198,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 4,
+   "execution_count": 6,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -289,6 +288,8 @@
     "    \"Qwen/Qwen2.5-0.5B-Instruct\",\n",
     "    \"Qwen/Qwen2.5-1.5B\",\n",
     "    \"Qwen/Qwen2.5-1.5B-Instruct\",\n",
+    "    \"Qwen/Qwen3-0.6B\",\n",
+    "    \"Qwen/Qwen3-1.7B\",\n",
     "    \"roneneldan/TinyStories-1Layer-21M\",\n",
     "    \"roneneldan/TinyStories-1M\",\n",
     "    \"roneneldan/TinyStories-28M\",\n",
@@ -323,7 +324,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 5,
+   "execution_count": 7,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -378,6 +379,8 @@
     "    \"Qwen/Qwen2.5-3B-Instruct\",\n",
     "    \"Qwen/Qwen2.5-7B\",\n",
     "    \"Qwen/Qwen2.5-7B-Instruct\",\n",
+    "    \"Qwen/Qwen3-4B\",\n",
+    "    \"Qwen/Qwen3-8B\",\n",
     "    \"stabilityai/stablelm-base-alpha-3b\",\n",
     "    \"stabilityai/stablelm-base-alpha-7b\",\n",
     "    \"stabilityai/stablelm-tuned-alpha-3b\",\n",
@@ -392,7 +395,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 6,
+   "execution_count": 8,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -425,7 +428,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 7,
+   "execution_count": 9,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -448,6 +451,7 @@
     "    \"Qwen/Qwen2.5-32B-Instruct\",\n",
     "    \"Qwen/Qwen2.5-72B\",\n",
     "    \"Qwen/Qwen2.5-72B-Instruct\",\n",
+    "    \"Qwen/Qwen3-14B\",\n",
     "    \"Qwen/QwQ-32B-Preview\",\n",
     "]\n",
     "\n",
@@ -456,7 +460,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 8,
+   "execution_count": 10,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -478,7 +482,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 9,
+   "execution_count": 11,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -496,7 +500,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 10,
+   "execution_count": 12,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -516,7 +520,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 11,
+   "execution_count": 13,
    "metadata": {},
    "outputs": [],
    "source": [
@@ -527,7 +531,7 @@
   },
   {
    "cell_type": "code",
-   "execution_count": 12,
+   "execution_count": 14,
    "metadata": {},
    "outputs": [
     {

docs/source/conf.py

@@ -107,6 +107,7 @@
 # https://www.sphinx-doc.org/en/master/usage/extensions/autodoc.html#confval-autodoc_default_options
 autodoc_default_options = {
     "exclude-members": ", ".join(functions_to_ignore),
+    "special-members": "__getitem__, __len__, __iter__",
 }