From f2eda961376b93161d07668fb0f22961516ede5e Mon Sep 17 00:00:00 2001
From: anzz1 <anzz1@live.com>
Date: Fri, 17 Mar 2023 03:58:34 +0200
Subject: [PATCH 01/26] Fix Makefile

---
 Makefile | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/Makefile b/Makefile
index cbbfb158913a..fb83d41badc5 100644
--- a/Makefile
+++ b/Makefile
@@ -176,7 +176,7 @@ $(info I CC:       $(CCV))
 $(info I CXX:      $(CXXV))
 $(info )
 
-default: main quantize
+default: chat quantize
 
 #
 # Build library
@@ -191,10 +191,6 @@ utils.o: utils.cpp utils.h
 clean:
 	rm -f *.o main quantize
 
-main: main.cpp ggml.o utils.o
-	$(CXX) $(CXXFLAGS) main.cpp ggml.o utils.o -o main $(LDFLAGS)
-	./main -h
-
 chat: chat.cpp ggml.o utils.o
 	$(CXX) $(CXXFLAGS) chat.cpp ggml.o utils.o -o chat $(LDFLAGS)
 

From 8a3d6f2ea4bf1f3e43876d65666bc777d18e2565 Mon Sep 17 00:00:00 2001
From: anzz1 <anzz1@live.com>
Date: Fri, 17 Mar 2023 04:00:40 +0200
Subject: [PATCH 02/26] CI: Enable Linux/MacOS builds

---
 .github/workflows/build.yml | 68 ++++++++++++++++++++-----------------
 1 file changed, 37 insertions(+), 31 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 59aac6314a33..034ab4de90bf 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -14,37 +14,43 @@ on:
     paths: ['CMakeLists.txt', 'Makefile', '**.h', '*.c', '**.cpp']
 
 jobs:
-#  ubuntu-latest:
-#    runs-on: ubuntu-latest
-#
-#    steps:
-#      - name: Clone
-#        uses: actions/checkout@v1
-#
-#      - name: Dependencies
-#        run: |
-#          sudo apt-get update
-#          sudo apt-get install build-essential
-#
-#      - name: Build
-#        run: |
-#          make
-#
-#  macOS-latest:
-#    runs-on: macOS-latest
-#
-#    steps:
-#      - name: Clone
-#        uses: actions/checkout@v1
-#
-#      - name: Dependencies
-#        run: |
-#          brew update
-#
-#      - name: Build
-#        run: |
-#          make
-#
+  ubuntu-latest:
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Clone
+        id: checkout
+        uses: actions/checkout@v1
+
+      - name: Dependencies
+        id: depends
+        run: |
+          sudo apt-get update
+          sudo apt-get install build-essential
+
+      - name: Build
+        id: make_build
+        run: |
+          make
+
+  macOS-latest:
+    runs-on: macOS-latest
+
+    steps:
+      - name: Clone
+        id: checkout
+        uses: actions/checkout@v1
+
+      - name: Dependencies
+        id: depends
+        run: |
+          brew update
+
+      - name: Build
+        id: make_build
+        run: |
+          make
+
   windows-latest:
     runs-on: windows-latest
 

From 5be098f51e6fbe4656d886b021a377d4c04310a0 Mon Sep 17 00:00:00 2001
From: Xiao-Yong Jin <jinxiaoyong@gmail.com>
Date: Fri, 17 Mar 2023 00:20:24 -0500
Subject: [PATCH 03/26] Compute remaining tokens along the way and exit if over

---
 chat.cpp | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)

diff --git a/chat.cpp b/chat.cpp
index 885d1f69a1e5..7bce57a2b10d 100644
--- a/chat.cpp
+++ b/chat.cpp
@@ -919,7 +919,8 @@ int main(int argc, char ** argv) {
                " - If you want to submit another line, end your input in '\\'.\n");
     }
 
-    int remaining_tokens = params.n_predict;
+    // we may want to slide the input window along with the context, but for now we restrict to the context length
+    int remaining_tokens = model.hparams.n_ctx - embd_inp.size();
     int input_consumed = 0;
     bool input_noecho = true;
 
@@ -935,7 +936,7 @@ int main(int argc, char ** argv) {
 
     
 
-    while (true) {
+    while (remaining_tokens > 0) {
         // predict
         if (embd.size() > 0) {
             const int64_t t_start_us = ggml_time_us();
@@ -980,7 +981,7 @@ int main(int argc, char ** argv) {
             input_noecho = false;
 
             // decrement remaining sampling budget
-            // --remaining_tokens;
+            --remaining_tokens;
         } else {
             // some user input remains from prompt or interaction, forward it to processing
             while (embd_inp.size() > input_consumed) {
@@ -1054,6 +1055,8 @@ int main(int argc, char ** argv) {
                     embd_inp.insert(embd_inp.end(), line_inp.begin(), line_inp.end());
                     embd_inp.insert(embd_inp.end(), response_inp.begin(), response_inp.end());
 
+                    remaining_tokens -= prompt_inp.size() + line_inp.size() + response_inp.size();
+
                     input_noecho = true; // do not echo this again
                 }
 

From af02c947989fffab4f87494533bb5b07e2919bc7 Mon Sep 17 00:00:00 2001
From: "Alex \"mcmonkey\" Goodwin" <git_commits@alexgoodwin.dev>
Date: Fri, 17 Mar 2023 04:49:41 -0700
Subject: [PATCH 04/26] add easy Windows install instructions to the readme

Also fix a typo of LLaMA's casing in the chat.cpp file, and add cmake's generated files to the gitignore
---
 .gitignore | 13 +++++++++++++
 README.md  | 23 ++++++++++++++++++++++-
 chat.cpp   |  2 +-
 3 files changed, 36 insertions(+), 2 deletions(-)

diff --git a/.gitignore b/.gitignore
index 5eb1ff1b873f..699f76c4ae91 100644
--- a/.gitignore
+++ b/.gitignore
@@ -15,9 +15,22 @@ build-sanitize-addr/
 build-sanitize-thread/
 
 models/*
+*.bin
 
 /main
 /quantize
 
 arm_neon.h
 compile_commands.json
+
+# Windows CMake files
+*.vcxproj
+*.filters
+*.cmake
+*.sln
+x64/
+Debug/
+Release/
+CMakeFiles/
+CMakeCache.txt
+*.dir/
diff --git a/README.md b/README.md
index 14b294f1a56b..55786bb9b39c 100644
--- a/README.md
+++ b/README.md
@@ -10,7 +10,7 @@ This combines the [LLaMA foundation model](https://github.com/facebookresearch/l
 
 ## Get started
 
-```
+```sh
 git clone https://github.com/antimatter15/alpaca.cpp
 cd alpaca.cpp
 
@@ -34,6 +34,27 @@ Save the `ggml-alpaca-7b-q4.bin` file in the same directory as your `./chat` exe
 
 The weights are based on the published fine-tunes from `alpaca-lora`, converted back into a pytorch checkpoint with a [modified script](https://github.com/tloen/alpaca-lora/pull/19) and then quantized with llama.cpp the regular way. 
 
+## Windows Setup
+
+- Download and install CMake: <https://cmake.org/download/>
+- Download and install `git`. If you've never used git before, consider a GUI client like <https://desktop.github.com/>
+- Clone this repo using your git client of choice (for GitHub Desktop, go to File -> Clone repository -> From URL and paste `https://github.com/antimatter15/alpaca.cpp` in as the URL)
+- Open a Windows Terminal inside the folder you cloned the repository to
+- Run the following commands one by one:
+
+```ps1
+cmake .
+cmake --build . --config Release
+```
+
+- Download the weights via any of the links in "Get started" above, and save the file as `ggml-alpaca-7b-q4.bin` in the main Alpaca directory.
+- In the terminal window, run this command:
+```ps1
+.\Release\chat.exe
+```
+- (You can add other launch options like `--n 8` as preferred onto the same line)
+- You can now type to the AI in the terminal and it will reply. Enjoy!
+
 ## Credit
 
 This combines [Facebook's LLaMA](https://github.com/facebookresearch/llama), [Stanford Alpaca](https://crfm.stanford.edu/2023/03/13/alpaca.html), [alpaca-lora](https://github.com/tloen/alpaca-lora) and [corresponding weights](https://huggingface.co/tloen/alpaca-lora-7b/tree/main) by Eric Wang (which uses [Jason Phang's implementation of LLaMA](https://github.com/huggingface/transformers/pull/21955) on top of Hugging Face Transformers), and [llama.cpp](https://github.com/ggerganov/llama.cpp) by Georgi Gerganov. The chat implementation is based on Matvey Soloviev's [Interactive Mode](https://github.com/ggerganov/llama.cpp/pull/61) for llama.cpp. Inspired by [Simon Willison's](https://til.simonwillison.net/llms/llama-7b-m2) getting started guide for LLaMA.
diff --git a/chat.cpp b/chat.cpp
index 885d1f69a1e5..5acb2bc95284 100644
--- a/chat.cpp
+++ b/chat.cpp
@@ -915,7 +915,7 @@ int main(int argc, char ** argv) {
 #if defined (__unix__) || (defined (__APPLE__) && defined (__MACH__)) || defined (_WIN32)
                " - Press Ctrl+C to interject at any time.\n"
 #endif
-               " - Press Return to return control to LLaMa.\n"
+               " - Press Return to return control to LLaMA.\n"
                " - If you want to submit another line, end your input in '\\'.\n");
     }
 

From e7bdee6cc9fb1d7a9d2cad5dd315333b3f0d9e51 Mon Sep 17 00:00:00 2001
From: anzz1 <anzz1@live.com>
Date: Fri, 17 Mar 2023 19:17:59 +0200
Subject: [PATCH 05/26] CI fine tuning

---
 .github/workflows/build.yml | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 034ab4de90bf..4387bafd47b8 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -13,6 +13,9 @@ on:
     types: [opened, synchronize, edited, reopened, review_requested, ready_for_review]
     paths: ['CMakeLists.txt', 'Makefile', '**.h', '*.c', '**.cpp']
 
+env:
+ BRANCH_NAME: ${{ github.head_ref || github.ref_name }}
+
 jobs:
   ubuntu-latest:
     runs-on: ubuntu-latest
@@ -76,7 +79,7 @@ jobs:
         id: pack_artifacts
         if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
         run: |
-          7z a alpaca-bin-win-x64-${{ steps.commit.outputs.short }}.zip .\build\Release\*
+          7z a alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-win-x64.zip .\build\Release\*
 
       - name: Create release
         id: create_release
@@ -85,7 +88,7 @@ jobs:
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         with:
-          tag_name: ${{ steps.commit.outputs.short }}
+          tag_name: ${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}
 
       - name: Upload release
         id: upload_release
@@ -95,8 +98,8 @@ jobs:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         with:
           upload_url: ${{ steps.create_release.outputs.upload_url }} 
-          asset_path: .\alpaca-bin-win-x64-${{ steps.commit.outputs.short }}.zip
-          asset_name: alpaca-bin-win-x64-${{ steps.commit.outputs.short }}.zip
+          asset_path: .\alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-win-x64.zip
+          asset_name: alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-win-x64.zip
           asset_content_type: application/octet-stream
 
 #  ubuntu-latest-gcc:

From d0f855a8df4a2ebe72d142cd343161b29a75f9ae Mon Sep 17 00:00:00 2001
From: Kevin Kwok <antimatter15@gmail.com>
Date: Fri, 17 Mar 2023 11:13:47 -0700
Subject: [PATCH 06/26] Update README.md

---
 README.md | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/README.md b/README.md
index 55786bb9b39c..0ca2e3812c85 100644
--- a/README.md
+++ b/README.md
@@ -55,6 +55,12 @@ cmake --build . --config Release
 - (You can add other launch options like `--n 8` as preferred onto the same line)
 - You can now type to the AI in the terminal and it will reply. Enjoy!
 
+## 13B
+
+TODO
+
+Torrent: `magnet:?xt=urn:btih:f3cf71b172129d6b5abccab393bc32253fac8159&dn=ggml-alpaca-13b-q4.bin&tr=udp%3A%2F%http://2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%https://t.co/zenhelfwRd%3A6969%2Fannounce&tr=https%3A%2F%https://t.co/zenhelfwRd%3A443%2Fannounce&tr=udp%3A%2F%https://t.co/RRAn1X65wE%3A6969%2Fannounce&tr=udp%3A%2F%https://t.co/uTXBeTLUMa%3A2810%2Fannounce`
+
 ## Credit
 
 This combines [Facebook's LLaMA](https://github.com/facebookresearch/llama), [Stanford Alpaca](https://crfm.stanford.edu/2023/03/13/alpaca.html), [alpaca-lora](https://github.com/tloen/alpaca-lora) and [corresponding weights](https://huggingface.co/tloen/alpaca-lora-7b/tree/main) by Eric Wang (which uses [Jason Phang's implementation of LLaMA](https://github.com/huggingface/transformers/pull/21955) on top of Hugging Face Transformers), and [llama.cpp](https://github.com/ggerganov/llama.cpp) by Georgi Gerganov. The chat implementation is based on Matvey Soloviev's [Interactive Mode](https://github.com/ggerganov/llama.cpp/pull/61) for llama.cpp. Inspired by [Simon Willison's](https://til.simonwillison.net/llms/llama-7b-m2) getting started guide for LLaMA.

From 7cd84a7027f4c60eb6fa7076333db47e35a169a9 Mon Sep 17 00:00:00 2001
From: Kevin Kwok <antimatter15@gmail.com>
Date: Fri, 17 Mar 2023 22:57:27 -0700
Subject: [PATCH 07/26] Update README.md

---
 README.md | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/README.md b/README.md
index 0ca2e3812c85..c5669646663c 100644
--- a/README.md
+++ b/README.md
@@ -59,7 +59,7 @@ cmake --build . --config Release
 
 TODO
 
-Torrent: `magnet:?xt=urn:btih:f3cf71b172129d6b5abccab393bc32253fac8159&dn=ggml-alpaca-13b-q4.bin&tr=udp%3A%2F%http://2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%https://t.co/zenhelfwRd%3A6969%2Fannounce&tr=https%3A%2F%https://t.co/zenhelfwRd%3A443%2Fannounce&tr=udp%3A%2F%https://t.co/RRAn1X65wE%3A6969%2Fannounce&tr=udp%3A%2F%https://t.co/uTXBeTLUMa%3A2810%2Fannounce`
+Torrent: `magnet:?xt=urn:btih:053b3d54d2e77ff020ebddf51dad681f2a651071&dn=ggml-alpaca-13b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce&tr=udp%3A%2F%2Ftracker.openbittorrent.com%3A6969%2Fannounce&tr=udp%3A%2F%2F9.rarbg.com%3A2810%2Fannounce`
 
 ## Credit
 

From 97d327e1bf6c9d166a38feeaf01d728a566e60b8 Mon Sep 17 00:00:00 2001
From: Kevin Kwok <antimatter15@gmail.com>
Date: Fri, 17 Mar 2023 23:43:09 -0700
Subject: [PATCH 08/26] Update chat.cpp

---
 chat.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/chat.cpp b/chat.cpp
index 5ee86ce78667..4bdd214ad7d5 100644
--- a/chat.cpp
+++ b/chat.cpp
@@ -30,7 +30,7 @@
 // determine number of model parts based on the dimension
 static const std::map<int, int> LLAMA_N_PARTS = {
     { 4096, 1 },
-    { 5120, 2 },
+    { 5120, 1 },
     { 6656, 4 },
     { 8192, 8 },
 };

From 96e0519ae1bd42412872607abacc1bc60e4df136 Mon Sep 17 00:00:00 2001
From: antimatter15 <antimatter15@gmail.com>
Date: Fri, 17 Mar 2023 23:46:31 -0700
Subject: [PATCH 09/26] extending context window

---
 README.md | 5 +++++
 chat.cpp  | 1 +
 2 files changed, 6 insertions(+)

diff --git a/README.md b/README.md
index c5669646663c..c8d6df9fa219 100644
--- a/README.md
+++ b/README.md
@@ -61,6 +61,11 @@ TODO
 
 Torrent: `magnet:?xt=urn:btih:053b3d54d2e77ff020ebddf51dad681f2a651071&dn=ggml-alpaca-13b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce&tr=udp%3A%2F%2Ftracker.openbittorrent.com%3A6969%2Fannounce&tr=udp%3A%2F%2F9.rarbg.com%3A2810%2Fannounce`
 
+
+```
+./chat -m ggml-alpaca-13b-q4.bin
+```
+
 ## Credit
 
 This combines [Facebook's LLaMA](https://github.com/facebookresearch/llama), [Stanford Alpaca](https://crfm.stanford.edu/2023/03/13/alpaca.html), [alpaca-lora](https://github.com/tloen/alpaca-lora) and [corresponding weights](https://huggingface.co/tloen/alpaca-lora-7b/tree/main) by Eric Wang (which uses [Jason Phang's implementation of LLaMA](https://github.com/huggingface/transformers/pull/21955) on top of Hugging Face Transformers), and [llama.cpp](https://github.com/ggerganov/llama.cpp) by Georgi Gerganov. The chat implementation is based on Matvey Soloviev's [Interactive Mode](https://github.com/ggerganov/llama.cpp/pull/61) for llama.cpp. Inspired by [Simon Willison's](https://til.simonwillison.net/llms/llama-7b-m2) getting started guide for LLaMA.
diff --git a/chat.cpp b/chat.cpp
index 4bdd214ad7d5..7ed8ea4c7b7c 100644
--- a/chat.cpp
+++ b/chat.cpp
@@ -798,6 +798,7 @@ int main(int argc, char ** argv) {
 
     params.temp = 0.1f;
     params.top_p = 0.95f;
+    params.n_ctx = 2048;
     params.interactive = true;
     params.interactive_start = true;
     params.use_color = true;

From e95e64bd49dcfbbdf880f5f94f42b13fa8437694 Mon Sep 17 00:00:00 2001
From: thement <40525767+thement@users.noreply.github.com>
Date: Fri, 17 Mar 2023 21:05:58 +0100
Subject: [PATCH 10/26] Implement non-greedy tokenizer that tries to maximize
 token lengths (#242)

* Implement non-greedy tokenizer that tries to maximize token lengths

* Insert single space in front of the prompt

- this is to match original llama tokenizer behavior

---------

Co-authored-by: Jakub Horak <jakub.horak@ibawizard.net>
---
 chat.cpp  |  2 ++
 utils.cpp | 68 ++++++++++++++++++++++++++++++++++---------------------
 2 files changed, 44 insertions(+), 26 deletions(-)

diff --git a/chat.cpp b/chat.cpp
index 7ed8ea4c7b7c..6d84af20d445 100644
--- a/chat.cpp
+++ b/chat.cpp
@@ -852,6 +852,8 @@ int main(int argc, char ** argv) {
 
     std::vector<float> logits;
 
+    // Add a space in front of the first character to match OG llama tokenizer behavior
+    params.prompt.insert(0, 1, ' ');
     // tokenize the prompt
     std::vector<gpt_vocab::id> embd_inp;// = ::llama_tokenize(vocab, params.prompt, true);
 
diff --git a/utils.cpp b/utils.cpp
index aa3ad1053da0..d739b5d48923 100644
--- a/utils.cpp
+++ b/utils.cpp
@@ -275,41 +275,57 @@ std::vector<gpt_vocab::id> gpt_tokenize(const gpt_vocab & vocab, const std::stri
     return tokens;
 }
 
+// TODO: Calculate this constant from the vocabulary
+#define MAX_TOKEN_LEN 18
+// SentencePiece implementation after https://guillaume-be.github.io/2020-05-30/sentence_piece
 std::vector<gpt_vocab::id> llama_tokenize(const gpt_vocab & vocab, const std::string & text, bool bos) {
-    //auto res = gpt_tokenize(vocab, text);
-
-    //if (bos) {
-    //    res.insert(res.begin(), 1); // TODO: replace with vocab.bos
-    //}
-
     std::vector<gpt_vocab::id> res;
-
-    if (bos) {
-        res.push_back(1); // TODO: replace with vocab.bos
-    }
-
-     //find the longest token that matches the text
-    int pos = 0;
-    while (true) {
-        int l = 0;
-        int t = 0;
-        for (const auto & kv : vocab.id_to_token) {
-            if (kv.second.size() < l) continue;
-            if (kv.second.size() > text.size() - pos) continue;
-            if (text.substr(pos, kv.second.size()) == kv.second) {
-                l = kv.second.size();
-                t = kv.first;
+    std::vector<int> score;
+    std::vector<gpt_vocab::id> prev;
+    int len = text.length();
+
+    score.resize(len + 1);
+    prev.resize(len + 1);
+
+    // Forward pass
+    for (int i = 0; i < len; i++) {
+        int max_len = std::min(len - i, MAX_TOKEN_LEN);
+        for (int sub_len = 1; sub_len <= len - i; sub_len++) {
+            auto sub = text.substr(i, sub_len);
+            auto token = vocab.token_to_id.find(sub);
+            if (token != vocab.token_to_id.end()) {
+                int token_score = sub.length() * sub.length();
+                int local_score = score[i] + token_score;
+                int next = i + sub_len;
+                if (score[next] < local_score) {
+                    score[next] = local_score;
+                    prev[next] = (*token).second;
+                }
             }
         }
+    }
 
-        if (l == 0) {
-            break;
+    // Backward pass
+    int i = len;
+    while (i > 0) {
+        gpt_vocab::id token_id = prev[i];
+        if (token_id == 0) {
+	    // TODO: Return error or something more meaningful
+            printf("failed to tokenize string!\n");
+	    break;
         }
+        res.push_back(token_id);
+        auto token = (*vocab.id_to_token.find(token_id)).second;
+        i -= token.length();
+    }
 
-        res.push_back(t);
-        pos += l;
+    if (bos) {
+        res.push_back(1); // TODO: replace with vocab.bos
     }
 
+    // Pieces are in reverse order so correct that
+    std::reverse(res.begin(), res.end());
+
     return res;
 }
 

From 1cb9215e5df342d26e399ec1405e5751c24ad572 Mon Sep 17 00:00:00 2001
From: antimatter15 <antimatter15@gmail.com>
Date: Sat, 18 Mar 2023 00:19:52 -0700
Subject: [PATCH 11/26] removing random prompt generation

---
 chat.cpp | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/chat.cpp b/chat.cpp
index 6d84af20d445..596a806d3bf2 100644
--- a/chat.cpp
+++ b/chat.cpp
@@ -815,9 +815,9 @@ int main(int argc, char ** argv) {
     fprintf(stderr, "%s: seed = %d\n", __func__, params.seed);
 
     std::mt19937 rng(params.seed);
-    if (params.prompt.empty()) {
-        params.prompt = gpt_random_prompt(rng);
-    }
+    // if (params.prompt.empty()) {
+    //     params.prompt = gpt_random_prompt(rng);
+    // }
 
 //    params.prompt = R"(// this function checks if the number n is prime
 //bool is_prime(int n) {)";
@@ -853,7 +853,7 @@ int main(int argc, char ** argv) {
     std::vector<float> logits;
 
     // Add a space in front of the first character to match OG llama tokenizer behavior
-    params.prompt.insert(0, 1, ' ');
+    // params.prompt.insert(0, 1, ' ');
     // tokenize the prompt
     std::vector<gpt_vocab::id> embd_inp;// = ::llama_tokenize(vocab, params.prompt, true);
 

From 501a8e19d98c6d442923035c7f275df71ee0a900 Mon Sep 17 00:00:00 2001
From: antimatter15 <antimatter15@gmail.com>
Date: Sat, 18 Mar 2023 00:28:05 -0700
Subject: [PATCH 12/26] adding to credit section

---
 README.md | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/README.md b/README.md
index c8d6df9fa219..65e537cbc741 100644
--- a/README.md
+++ b/README.md
@@ -57,7 +57,7 @@ cmake --build . --config Release
 
 ## 13B
 
-TODO
+TODO: write more docs here (PRs welcome)
 
 Torrent: `magnet:?xt=urn:btih:053b3d54d2e77ff020ebddf51dad681f2a651071&dn=ggml-alpaca-13b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce&tr=udp%3A%2F%2Ftracker.openbittorrent.com%3A6969%2Fannounce&tr=udp%3A%2F%2F9.rarbg.com%3A2810%2Fannounce`
 
@@ -68,7 +68,7 @@ Torrent: `magnet:?xt=urn:btih:053b3d54d2e77ff020ebddf51dad681f2a651071&dn=ggml-a
 
 ## Credit
 
-This combines [Facebook's LLaMA](https://github.com/facebookresearch/llama), [Stanford Alpaca](https://crfm.stanford.edu/2023/03/13/alpaca.html), [alpaca-lora](https://github.com/tloen/alpaca-lora) and [corresponding weights](https://huggingface.co/tloen/alpaca-lora-7b/tree/main) by Eric Wang (which uses [Jason Phang's implementation of LLaMA](https://github.com/huggingface/transformers/pull/21955) on top of Hugging Face Transformers), and [llama.cpp](https://github.com/ggerganov/llama.cpp) by Georgi Gerganov. The chat implementation is based on Matvey Soloviev's [Interactive Mode](https://github.com/ggerganov/llama.cpp/pull/61) for llama.cpp. Inspired by [Simon Willison's](https://til.simonwillison.net/llms/llama-7b-m2) getting started guide for LLaMA.
+This combines [Facebook's LLaMA](https://github.com/facebookresearch/llama), [Stanford Alpaca](https://crfm.stanford.edu/2023/03/13/alpaca.html), [alpaca-lora](https://github.com/tloen/alpaca-lora) and [corresponding weights](https://huggingface.co/tloen/alpaca-lora-7b/tree/main) by Eric Wang (which uses [Jason Phang's implementation of LLaMA](https://github.com/huggingface/transformers/pull/21955) on top of Hugging Face Transformers), and [llama.cpp](https://github.com/ggerganov/llama.cpp) by Georgi Gerganov. The chat implementation is based on Matvey Soloviev's [Interactive Mode](https://github.com/ggerganov/llama.cpp/pull/61) for llama.cpp. Inspired by [Simon Willison's](https://til.simonwillison.net/llms/llama-7b-m2) getting started guide for LLaMA. [Andy Matuschak](https://twitter.com/andy_matuschak/status/1636769182066053120)'s thread on adapting this to 13B, using fine tuning weights by [Sam Witteveen](https://huggingface.co/samwit/alpaca13B-lora). 
 
 
 ## Disclaimer

From 7e126618c41a12f2ccd8a7ad2ba2c14274aa60cd Mon Sep 17 00:00:00 2001
From: antimatter15 <antimatter15@gmail.com>
Date: Sat, 18 Mar 2023 00:34:53 -0700
Subject: [PATCH 13/26] ci releases for mac and linux

---
 .github/workflows/build.yml | 36 ++++++++++++++++++++++++++++++++++++
 1 file changed, 36 insertions(+)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 4387bafd47b8..9f9a58c81c38 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -36,6 +36,24 @@ jobs:
         run: |
           make
 
+      - name: Pack artifacts
+        id: pack_artifacts
+        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+        run: |
+          zip alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-linux-x64.zip chat
+
+      - name: Upload release
+        id: upload_release
+        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+        uses: actions/upload-release-asset@v1
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        with:
+          upload_url: ${{ steps.create_release.outputs.upload_url }} 
+          asset_path: .\alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-linux-x64.zip
+          asset_name: alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-linux-x64.zip
+          asset_content_type: application/octet-stream
+
   macOS-latest:
     runs-on: macOS-latest
 
@@ -54,6 +72,24 @@ jobs:
         run: |
           make
 
+      - name: Pack artifacts
+        id: pack_artifacts
+        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+        run: |
+          zip alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-mac-x64.zip chat
+
+      - name: Upload release
+        id: upload_release
+        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+        uses: actions/upload-release-asset@v1
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        with:
+          upload_url: ${{ steps.create_release.outputs.upload_url }} 
+          asset_path: .\alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-mac-x64.zip
+          asset_name: alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-mac-x64.zip
+          asset_content_type: application/octet-stream
+
   windows-latest:
     runs-on: windows-latest
 

From 1c62e3598419e23ec91664cce66d8dcdc806eb44 Mon Sep 17 00:00:00 2001
From: antimatter15 <antimatter15@gmail.com>
Date: Sat, 18 Mar 2023 00:37:34 -0700
Subject: [PATCH 14/26] create release

---
 .github/workflows/build.yml | 19 +++++++++++++++++++
 1 file changed, 19 insertions(+)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 9f9a58c81c38..bff3e5609906 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -41,6 +41,15 @@ jobs:
         if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
         run: |
           zip alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-linux-x64.zip chat
+      
+      - name: Create release
+        id: create_release
+        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+        uses: zendesk/action-create-release@v1
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        with:
+          tag_name: ${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}
 
       - name: Upload release
         id: upload_release
@@ -72,11 +81,21 @@ jobs:
         run: |
           make
 
+
       - name: Pack artifacts
         id: pack_artifacts
         if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
         run: |
           zip alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-mac-x64.zip chat
+      
+      - name: Create release
+        id: create_release
+        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+        uses: zendesk/action-create-release@v1
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        with:
+          tag_name: ${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}
 
       - name: Upload release
         id: upload_release

From 3f7d187b6b3338bc32d4f60029ab40fd38dbeeb5 Mon Sep 17 00:00:00 2001
From: antimatter15 <antimatter15@gmail.com>
Date: Sat, 18 Mar 2023 00:55:00 -0700
Subject: [PATCH 15/26] more copying stuff

---
 .github/workflows/build.yml | 9 +++++++++
 1 file changed, 9 insertions(+)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index bff3e5609906..77b2586c4779 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -36,6 +36,11 @@ jobs:
         run: |
           make
 
+      - name: Set commit hash variables
+        id: commit
+        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+        uses: pr-mpt/actions-commit-hash@v2
+        
       - name: Pack artifacts
         id: pack_artifacts
         if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
@@ -81,6 +86,10 @@ jobs:
         run: |
           make
 
+      - name: Set commit hash variables
+        id: commit
+        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+        uses: pr-mpt/actions-commit-hash@v2
 
       - name: Pack artifacts
         id: pack_artifacts

From 564b861bac1971c20cd0dfff9568787d9e6b31e6 Mon Sep 17 00:00:00 2001
From: antimatter15 <antimatter15@gmail.com>
Date: Sat, 18 Mar 2023 01:06:44 -0700
Subject: [PATCH 16/26] archiving artifacts

---
 .github/workflows/build.yml | 88 ++++++++++++++-----------------------
 1 file changed, 32 insertions(+), 56 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 77b2586c4779..4aa2bdef5b2a 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -36,37 +36,13 @@ jobs:
         run: |
           make
 
-      - name: Set commit hash variables
-        id: commit
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
-        uses: pr-mpt/actions-commit-hash@v2
-        
-      - name: Pack artifacts
-        id: pack_artifacts
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
-        run: |
-          zip alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-linux-x64.zip chat
-      
-      - name: Create release
-        id: create_release
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
-        uses: zendesk/action-create-release@v1
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+      - name: Archive production artifacts
+        uses: actions/upload-artifact@v3
         with:
-          tag_name: ${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}
+          name: ubuntu
+          path: |
+            chat
 
-      - name: Upload release
-        id: upload_release
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
-        uses: actions/upload-release-asset@v1
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        with:
-          upload_url: ${{ steps.create_release.outputs.upload_url }} 
-          asset_path: .\alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-linux-x64.zip
-          asset_name: alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-linux-x64.zip
-          asset_content_type: application/octet-stream
 
   macOS-latest:
     runs-on: macOS-latest
@@ -86,37 +62,37 @@ jobs:
         run: |
           make
 
-      - name: Set commit hash variables
-        id: commit
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
-        uses: pr-mpt/actions-commit-hash@v2
+      - name: Archive production artifacts
+        uses: actions/upload-artifact@v3
+        with:
+          name: macos
+          path: |
+            chat
 
-      - name: Pack artifacts
-        id: pack_artifacts
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+  macOS-arm64:
+    runs-on: macOS-arm64
+
+    steps:
+      - name: Clone
+        id: checkout
+        uses: actions/checkout@v1
+
+      - name: Dependencies
+        id: depends
         run: |
-          zip alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-mac-x64.zip chat
-      
-      - name: Create release
-        id: create_release
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
-        uses: zendesk/action-create-release@v1
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-        with:
-          tag_name: ${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}
+          brew update
 
-      - name: Upload release
-        id: upload_release
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
-        uses: actions/upload-release-asset@v1
-        env:
-          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+      - name: Build
+        id: make_build
+        run: |
+          make
+
+      - name: Archive production artifacts
+        uses: actions/upload-artifact@v3
         with:
-          upload_url: ${{ steps.create_release.outputs.upload_url }} 
-          asset_path: .\alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-mac-x64.zip
-          asset_name: alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-mac-x64.zip
-          asset_content_type: application/octet-stream
+          name: macos
+          path: |
+            chat
 
   windows-latest:
     runs-on: windows-latest

From ddc4e24cb85fb4a05d9411e24b344974fb9f8cf4 Mon Sep 17 00:00:00 2001
From: antimatter15 <antimatter15@gmail.com>
Date: Sat, 18 Mar 2023 01:11:02 -0700
Subject: [PATCH 17/26] maybe macos-arm64 is case sensitive

---
 .github/workflows/build.yml | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 4aa2bdef5b2a..48ec93b9bc2d 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -69,8 +69,8 @@ jobs:
           path: |
             chat
 
-  macOS-arm64:
-    runs-on: macOS-arm64
+  macos-arm64:
+    runs-on: macos-arm64
 
     steps:
       - name: Clone

From 4a524c51ba42254fdae63abcf26d38f64dd4fc28 Mon Sep 17 00:00:00 2001
From: antimatter15 <antimatter15@gmail.com>
Date: Sat, 18 Mar 2023 01:15:51 -0700
Subject: [PATCH 18/26] commenting out aarch

---
 .github/workflows/build.yml | 40 ++++++++++++++++++-------------------
 1 file changed, 20 insertions(+), 20 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index 48ec93b9bc2d..d206f21bbeaf 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -69,30 +69,30 @@ jobs:
           path: |
             chat
 
-  macos-arm64:
-    runs-on: macos-arm64
+  # macos-arm64:
+  #   runs-on: macos-arm64
 
-    steps:
-      - name: Clone
-        id: checkout
-        uses: actions/checkout@v1
+  #   steps:
+  #     - name: Clone
+  #       id: checkout
+  #       uses: actions/checkout@v1
 
-      - name: Dependencies
-        id: depends
-        run: |
-          brew update
+  #     - name: Dependencies
+  #       id: depends
+  #       run: |
+  #         brew update
 
-      - name: Build
-        id: make_build
-        run: |
-          make
+  #     - name: Build
+  #       id: make_build
+  #       run: |
+  #         make
 
-      - name: Archive production artifacts
-        uses: actions/upload-artifact@v3
-        with:
-          name: macos
-          path: |
-            chat
+  #     - name: Archive production artifacts
+  #       uses: actions/upload-artifact@v3
+  #       with:
+  #         name: macos
+  #         path: |
+  #           chat
 
   windows-latest:
     runs-on: windows-latest

From a83e2e7a2447d782ed41810535a94d4c73b3dc8f Mon Sep 17 00:00:00 2001
From: Rupesh Sreeraman <exmplayer.dev@gmail.com>
Date: Sat, 18 Mar 2023 16:41:20 +0530
Subject: [PATCH 19/26] Windows console ANSI color issue fixed

---
 chat.cpp | 6 ++++++
 1 file changed, 6 insertions(+)

diff --git a/chat.cpp b/chat.cpp
index 596a806d3bf2..f2ccf12fcbce 100644
--- a/chat.cpp
+++ b/chat.cpp
@@ -16,6 +16,7 @@
 #include <unistd.h>
 #elif defined (_WIN32)
 #include <signal.h>
+#include <Windows.h>
 #endif
 
 #define ANSI_COLOR_RED     "\x1b[31m"
@@ -886,6 +887,11 @@ int main(int argc, char ** argv) {
         sigaction(SIGINT, &sigint_action, NULL);
 #elif defined (_WIN32)
         signal(SIGINT, sigint_handler);
+		//Windows console ANSI color fix 
+		HANDLE hConsole = GetStdHandle(STD_OUTPUT_HANDLE);
+		DWORD mode;
+		GetConsoleMode(hConsole, &mode);
+		SetConsoleMode(hConsole, mode | ENABLE_VIRTUAL_TERMINAL_PROCESSING);
 #endif
 
         fprintf(stderr, "%s: interactive mode on.\n", __func__);

From bb60fdaf32328651e049ff22a7ab43fb6575bbc1 Mon Sep 17 00:00:00 2001
From: Marius Ciocanel <marius.ciocanel@gmail.com>
Date: Sat, 18 Mar 2023 13:10:25 +0000
Subject: [PATCH 20/26] Update command for downloading the weights to use
 `curl`

`curl` is preinstalled on macOS and the new command is equivalent to the `wget` version but avoids having to install wget
---
 README.md | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/README.md b/README.md
index 65e537cbc741..d7358dad076f 100644
--- a/README.md
+++ b/README.md
@@ -25,9 +25,9 @@ Alternatively you can download them with IPFS.
 
 ```
 # any of these will work
-wget -O ggml-alpaca-7b-q4.bin -c https://gateway.estuary.tech/gw/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
-wget -O ggml-alpaca-7b-q4.bin -c https://ipfs.io/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
-wget -O ggml-alpaca-7b-q4.bin -c https://cloudflare-ipfs.com/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
+curl -o ggml-alpaca-7b-q4.bin -C - https://gateway.estuary.tech/gw/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
+curl -o ggml-alpaca-7b-q4.bin -C - https://ipfs.io/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
+curl -o ggml-alpaca-7b-q4.bin -C - https://cloudflare-ipfs.com/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
 ```
 
 Save the `ggml-alpaca-7b-q4.bin` file in the same directory as your `./chat` executable. 

From c0e1cb53c7220f673b3ae1b8a3927fcb906c22af Mon Sep 17 00:00:00 2001
From: Nato Boram <nato.boram@gmail.com>
Date: Sat, 18 Mar 2023 14:45:10 -0400
Subject: [PATCH 21/26] =?UTF-8?q?=F0=9F=99=88=20Add=20output=20`chat`=20to?=
 =?UTF-8?q?=20`.gitignore`?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 .gitignore | 2 ++
 1 file changed, 2 insertions(+)

diff --git a/.gitignore b/.gitignore
index 699f76c4ae91..da9e6bed8008 100644
--- a/.gitignore
+++ b/.gitignore
@@ -1,3 +1,5 @@
+/chat
+
 *.o
 *.a
 .cache/

From f69062f68e7bffb6060ca9b9d778c14fa889b863 Mon Sep 17 00:00:00 2001
From: anzz1 <anzz1@live.com>
Date: Sat, 18 Mar 2023 21:51:12 +0200
Subject: [PATCH 22/26] Do the windows ANSI color fix properly

---
 chat.cpp | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)

diff --git a/chat.cpp b/chat.cpp
index f2ccf12fcbce..6524813727bd 100644
--- a/chat.cpp
+++ b/chat.cpp
@@ -887,11 +887,12 @@ int main(int argc, char ** argv) {
         sigaction(SIGINT, &sigint_action, NULL);
 #elif defined (_WIN32)
         signal(SIGINT, sigint_handler);
-		//Windows console ANSI color fix 
-		HANDLE hConsole = GetStdHandle(STD_OUTPUT_HANDLE);
-		DWORD mode;
-		GetConsoleMode(hConsole, &mode);
-		SetConsoleMode(hConsole, mode | ENABLE_VIRTUAL_TERMINAL_PROCESSING);
+
+        // Windows console ANSI color fix
+        DWORD mode;
+        HANDLE hConsole = GetStdHandle(STD_OUTPUT_HANDLE);
+        if (hConsole && hConsole != INVALID_HANDLE_VALUE && GetConsoleMode(hConsole, &mode))
+            SetConsoleMode(hConsole, mode | ENABLE_VIRTUAL_TERMINAL_PROCESSING);
 #endif
 
         fprintf(stderr, "%s: interactive mode on.\n", __func__);

From 1b19586681c371844492b26aa4bcf1f157bf863f Mon Sep 17 00:00:00 2001
From: anzz1 <anzz1@live.com>
Date: Sat, 18 Mar 2023 22:21:58 +0200
Subject: [PATCH 23/26] Init the var too

---
 chat.cpp | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/chat.cpp b/chat.cpp
index 6524813727bd..38b39771ad98 100644
--- a/chat.cpp
+++ b/chat.cpp
@@ -889,7 +889,7 @@ int main(int argc, char ** argv) {
         signal(SIGINT, sigint_handler);
 
         // Windows console ANSI color fix
-        DWORD mode;
+        DWORD mode = 0;
         HANDLE hConsole = GetStdHandle(STD_OUTPUT_HANDLE);
         if (hConsole && hConsole != INVALID_HANDLE_VALUE && GetConsoleMode(hConsole, &mode))
             SetConsoleMode(hConsole, mode | ENABLE_VIRTUAL_TERMINAL_PROCESSING);

From e83984f583ec865ed445cf08ca9e96146aaa4ac7 Mon Sep 17 00:00:00 2001
From: anzz1 <anzz1@live.com>
Date: Mon, 20 Mar 2023 02:31:42 +0200
Subject: [PATCH 24/26] add cached torrent links

for people having trouble with magnet links, as torrents always work
---
 README.md | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)

diff --git a/README.md b/README.md
index d7358dad076f..a1ecbe78cef8 100644
--- a/README.md
+++ b/README.md
@@ -18,7 +18,11 @@ make chat
 ./chat
 ```
 
-You can download the weights for `ggml-alpaca-7b-q4.bin` with BitTorrent `magnet:?xt=urn:btih:5aaceaec63b03e51a98f04fd5c42320b2a033010&dn=ggml-alpaca-7b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce`
+You can download the weights for `ggml-alpaca-7b-q4.bin` with BitTorrent:
+
+magnet: `magnet:?xt=urn:btih:5aaceaec63b03e51a98f04fd5c42320b2a033010&dn=ggml-alpaca-7b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce`  
+torrent: https://btcache.me/torrent/5AACEAEC63B03E51A98F04FD5C42320B2A033010  
+torrent: https://torrage.info/torrent.php?h=5aaceaec63b03e51a98f04fd5c42320b2a033010  
 
 
 Alternatively you can download them with IPFS.
@@ -59,7 +63,11 @@ cmake --build . --config Release
 
 TODO: write more docs here (PRs welcome)
 
-Torrent: `magnet:?xt=urn:btih:053b3d54d2e77ff020ebddf51dad681f2a651071&dn=ggml-alpaca-13b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce&tr=udp%3A%2F%2Ftracker.openbittorrent.com%3A6969%2Fannounce&tr=udp%3A%2F%2F9.rarbg.com%3A2810%2Fannounce`
+You can download the weights for `ggml-alpaca-13b-q4.bin` with BitTorrent:
+
+magnet: `magnet:?xt=urn:btih:053b3d54d2e77ff020ebddf51dad681f2a651071&dn=ggml-alpaca-13b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce&tr=udp%3A%2F%2Ftracker.openbittorrent.com%3A6969%2Fannounce&tr=udp%3A%2F%2F9.rarbg.com%3A2810%2Fannounce`  
+torrent: https://btcache.me/torrent/053B3D54D2E77FF020EBDDF51DAD681F2A651071  
+torrent: https://torrage.info/torrent.php?h=053b3d54d2e77ff020ebddf51dad681f2a651071
 
 
 ```

From 99f3908c515c73c7acbda307565234f33ec9738d Mon Sep 17 00:00:00 2001
From: Kevin Kwok <antimatter15@gmail.com>
Date: Mon, 20 Mar 2023 03:41:11 -0700
Subject: [PATCH 25/26] Automatically Generate Releases for Mac, Linux, Windows
 (#81)

* trying to build for arm

* cross compiling?

* only mac

* gh actions

* fix

* fixing dash

* ensure that its executable

* adding back windows

* unindent

* typo

* oops

* runs on

* more release stuff

* more release stuff

* fix workflow

* adding back linux builds

* conditional stuff
---
 .github/workflows/build.yml | 322 ++++++++++--------------------------
 Makefile                    |   7 +
 2 files changed, 94 insertions(+), 235 deletions(-)

diff --git a/.github/workflows/build.yml b/.github/workflows/build.yml
index d206f21bbeaf..d185e1ac5c59 100644
--- a/.github/workflows/build.yml
+++ b/.github/workflows/build.yml
@@ -36,15 +36,18 @@ jobs:
         run: |
           make
 
+      - name: Zip executables
+        run: zip alpaca-linux.zip chat
+
       - name: Archive production artifacts
         uses: actions/upload-artifact@v3
         with:
-          name: ubuntu
+          name: linux
           path: |
-            chat
+            alpaca-linux.zip
 
 
-  macOS-latest:
+  macos-latest:
     runs-on: macOS-latest
 
     steps:
@@ -60,39 +63,45 @@ jobs:
       - name: Build
         id: make_build
         run: |
-          make
+          make chat_mac
+
+      - name: Codesign executable
+        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+        env: 
+          MACOS_CERTIFICATE: ${{ secrets.APPLE_DEVELOPER_CERTIFICATE_P12_BASE64 }}
+          MACOS_CERTIFICATE_PWD: ${{ secrets.APPLE_DEVELOPER_CERTIFICATE_PASSWORD }}
+          KEYCHAIN_PWD: ${{ secrets.KEYCHAIN_PASSWORD }}
+          MACOS_CERT_ID: ${{ secrets.APPLE_DEVELOPER_CERTIFICATE_ID }}
+
+        run: |
+          echo $MACOS_CERTIFICATE | base64 --decode > certificate.p12
+          security create-keychain -p $KEYCHAIN_PWD build.keychain
+          security default-keychain -s build.keychain
+          security unlock-keychain -p $KEYCHAIN_PWD build.keychain
+          security import certificate.p12 -k build.keychain -P $MACOS_CERTIFICATE_PWD -T /usr/bin/codesign
+          security set-key-partition-list -S apple-tool:,apple:,codesign: -s -k $KEYCHAIN_PWD build.keychain
+          /usr/bin/codesign --options runtime --force -s $MACOS_CERT_ID ./chat_mac -v
+
+      - name: Zip executables
+        run: zip alpaca-mac.zip chat_mac
+
+      - name: Notarize executables
+        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+        env: 
+          PROD_MACOS_NOTARIZATION_APPLE_ID: ${{ secrets.AC_USERNAME }}
+          PROD_MACOS_NOTARIZATION_PWD: ${{ secrets.AC_PASSWORD }}
+          PROD_MACOS_NOTARIZATION_TEAM_ID: ${{ secrets.PROD_MACOS_NOTARIZATION_TEAM_ID }}
+        run: |
+          xcrun notarytool store-credentials "notarytool-profile" --apple-id "$PROD_MACOS_NOTARIZATION_APPLE_ID" --team-id "$PROD_MACOS_NOTARIZATION_TEAM_ID" --password "$PROD_MACOS_NOTARIZATION_PWD"
+          xcrun notarytool submit "alpaca-mac.zip" --keychain-profile "notarytool-profile" --wait
 
       - name: Archive production artifacts
         uses: actions/upload-artifact@v3
         with:
           name: macos
           path: |
-            chat
-
-  # macos-arm64:
-  #   runs-on: macos-arm64
-
-  #   steps:
-  #     - name: Clone
-  #       id: checkout
-  #       uses: actions/checkout@v1
-
-  #     - name: Dependencies
-  #       id: depends
-  #       run: |
-  #         brew update
-
-  #     - name: Build
-  #       id: make_build
-  #       run: |
-  #         make
+            alpaca-mac.zip
 
-  #     - name: Archive production artifacts
-  #       uses: actions/upload-artifact@v3
-  #       with:
-  #         name: macos
-  #         path: |
-  #           chat
 
   windows-latest:
     runs-on: windows-latest
@@ -110,227 +119,70 @@ jobs:
           cmake ..
           cmake --build . --config Release
 
-      - name: Set commit hash variables
-        id: commit
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
-        uses: pr-mpt/actions-commit-hash@v2
-
       - name: Pack artifacts
         id: pack_artifacts
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
         run: |
-          7z a alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-win-x64.zip .\build\Release\*
+          7z a alpaca-win.zip .\build\Release\*
+
+      - name: Archive production artifacts
+        uses: actions/upload-artifact@v3
+        with:
+          name: windows
+          path: |
+            alpaca-win.zip
+
+
+  release:
+    runs-on: ubuntu-latest
+    if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+    needs: 
+      - windows-latest
+      - macos-latest
+      - ubuntu-latest
+
+    steps:
+      - name: Download all workflow run artifacts
+        uses: actions/download-artifact@v3
+
+      - name: Set commit hash variables
+        id: commit
+        uses: pr-mpt/actions-commit-hash@v2
 
       - name: Create release
         id: create_release
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
         uses: zendesk/action-create-release@v1
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         with:
-          tag_name: ${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}
+          tag_name: ${{ steps.commit.outputs.short }}
 
-      - name: Upload release
-        id: upload_release
-        if: ${{ ( github.event_name == 'push' && github.ref == 'refs/heads/master' ) || github.event.inputs.create_release == 'true' }}
+      - name: Upload windows release
         uses: actions/upload-release-asset@v1
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
         with:
           upload_url: ${{ steps.create_release.outputs.upload_url }} 
-          asset_path: .\alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-win-x64.zip
-          asset_name: alpaca-${{ env.BRANCH_NAME }}-${{ steps.commit.outputs.short }}-bin-win-x64.zip
+          asset_path: windows/alpaca-win.zip
+          asset_name: alpaca-win.zip
           asset_content_type: application/octet-stream
 
-#  ubuntu-latest-gcc:
-#    runs-on: ubuntu-latest
-#
-#    strategy:
-#      matrix:
-#        build: [Debug, Release]
-#
-#    steps:
-#      - name: Clone
-#        uses: actions/checkout@v1
-#
-#      - name: Dependencies
-#        run: |
-#          sudo apt-get update
-#          sudo apt-get install build-essential
-#          sudo apt-get install cmake
-#
-#      - name: Configure
-#        run: cmake . -DCMAKE_BUILD_TYPE=${{ matrix.build }}
-#
-#      - name: Build
-#        run: |
-#          make
-#
-#  ubuntu-latest-clang:
-#    runs-on: ubuntu-latest
-#
-#    strategy:
-#      matrix:
-#        build: [Debug, Release]
-#
-#    steps:
-#      - name: Clone
-#        uses: actions/checkout@v1
-#
-#      - name: Dependencies
-#        run: |
-#          sudo apt-get update
-#          sudo apt-get install build-essential
-#          sudo apt-get install cmake
-#
-#      - name: Configure
-#        run: cmake . -DCMAKE_BUILD_TYPE=${{ matrix.build }} -DCMAKE_CXX_COMPILER=clang++ -DCMAKE_C_COMPILER=clang
-#
-#      - name: Build
-#        run: |
-#          make
-#
-#  ubuntu-latest-gcc-sanitized:
-#    runs-on: ubuntu-latest
-#
-#    strategy:
-#      matrix:
-#        sanitizer: [ADDRESS, THREAD, UNDEFINED]
-#
-#    steps:
-#      - name: Clone
-#        uses: actions/checkout@v1
-#
-#      - name: Dependencies
-#        run: |
-#          sudo apt-get update
-#          sudo apt-get install build-essential
-#          sudo apt-get install cmake
-#
-#      - name: Configure
-#        run: cmake . -DCMAKE_BUILD_TYPE=Debug -DLLAMA_SANITIZE_${{ matrix.sanitizer }}=ON
-#
-#      - name: Build
-#        run: |
-#          make
-#
-#  windows:
-#    runs-on: windows-latest
-#
-#    strategy:
-#      matrix:
-#        build: [Release]
-#        arch: [Win32, x64]
-#        include:
-#          - arch: Win32
-#            s2arc: x86
-#          - arch: x64
-#            s2arc: x64
-#
-#    steps:
-#      - name: Clone
-#        uses: actions/checkout@v1
-#
-#      - name: Add msbuild to PATH
-#        uses: microsoft/setup-msbuild@v1
-#
-#      - name: Configure
-#        run: >
-#          cmake -S . -B ./build -A ${{ matrix.arch }}
-#          -DCMAKE_BUILD_TYPE=${{ matrix.build }}
-#
-#      - name: Build
-#        run: |
-#          cd ./build
-#          msbuild ALL_BUILD.vcxproj -t:build -p:configuration=${{ matrix.build }} -p:platform=${{ matrix.arch }}
-#
-#      - name: Upload binaries
-#        uses: actions/upload-artifact@v1
-#        with:
-#          name: llama-bin-${{ matrix.arch }}
-#          path: build/bin/${{ matrix.build }}
-#
-#  windows-blas:
-#    runs-on: windows-latest
-#
-#    strategy:
-#      matrix:
-#        build: [Release]
-#        arch: [Win32, x64]
-#        blas: [ON]
-#        include:
-#          - arch: Win32
-#            obzip: https://github.com/xianyi/OpenBLAS/releases/download/v0.3.21/OpenBLAS-0.3.21-x86.zip
-#            s2arc: x86
-#          - arch: x64
-#            obzip: https://github.com/xianyi/OpenBLAS/releases/download/v0.3.21/OpenBLAS-0.3.21-x64.zip
-#            s2arc: x64
-#
-#    steps:
-#      - name: Clone
-#        uses: actions/checkout@v1
-#
-#      - name: Add msbuild to PATH
-#        uses: microsoft/setup-msbuild@v1
-#
-#      - name: Fetch OpenBLAS
-#        if: matrix.blas == 'ON'
-#        run: |
-#          C:/msys64/usr/bin/wget.exe -qO blas.zip ${{ matrix.obzip }}
-#          7z x blas.zip -oblas -y
-#          copy blas/include/cblas.h .
-#          copy blas/include/openblas_config.h .
-#          echo "blasdir=$env:GITHUB_WORKSPACE/blas" >> $env:GITHUB_ENV
-#
-#      - name: Configure
-#        run: >
-#          cmake -S . -B ./build -A ${{ matrix.arch }}
-#          -DCMAKE_BUILD_TYPE=${{ matrix.build }}
-#          -DLLAMA_SUPPORT_OPENBLAS=${{ matrix.blas }}
-#          -DCMAKE_LIBRARY_PATH="$env:blasdir/lib"
-#
-#      - name: Build
-#        run: |
-#          cd ./build
-#          msbuild ALL_BUILD.vcxproj -t:build -p:configuration=${{ matrix.build }} -p:platform=${{ matrix.arch }}
-#
-#      - name: Copy libopenblas.dll
-#        if: matrix.blas == 'ON'
-#        run: copy "$env:blasdir/bin/libopenblas.dll" build/bin/${{ matrix.build }}
-#
-#      - name: Upload binaries
-#        if: matrix.blas == 'ON'
-#        uses: actions/upload-artifact@v1
-#        with:
-#          name: llama-blas-bin-${{ matrix.arch }}
-#          path: build/bin/${{ matrix.build }}
-#
-#  emscripten:
-#    runs-on: ubuntu-latest
-#
-#    strategy:
-#      matrix:
-#        build: [Release]
-#
-#    steps:
-#      - name: Clone
-#        uses: actions/checkout@v1
-#
-#      - name: Dependencies
-#        run: |
-#          wget -q https://github.com/emscripten-core/emsdk/archive/master.tar.gz
-#          tar -xvf master.tar.gz
-#          emsdk-master/emsdk update
-#          emsdk-master/emsdk install latest
-#          emsdk-master/emsdk activate latest
-#
-#      - name: Configure
-#        run: echo "tmp"
-#
-#      - name: Build
-#        run: |
-#          pushd emsdk-master
-#          source ./emsdk_env.sh
-#          popd
-#          emcmake cmake . -DCMAKE_BUILD_TYPE=${{ matrix.build }}
-#          make
+      - name: Upload mac release
+        uses: actions/upload-release-asset@v1
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        with:
+          upload_url: ${{ steps.create_release.outputs.upload_url }} 
+          asset_path: macos/alpaca-mac.zip
+          asset_name: alpaca-mac.zip
+          asset_content_type: application/octet-stream
+
+      - name: Upload linux release
+        
+        uses: actions/upload-release-asset@v1
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+        with:
+          upload_url: ${{ steps.create_release.outputs.upload_url }} 
+          asset_path: linux/alpaca-linux.zip
+          asset_name: alpaca-linux.zip
+          asset_content_type: application/octet-stream
diff --git a/Makefile b/Makefile
index fb83d41badc5..93da626f8836 100644
--- a/Makefile
+++ b/Makefile
@@ -194,6 +194,13 @@ clean:
 chat: chat.cpp ggml.o utils.o
 	$(CXX) $(CXXFLAGS) chat.cpp ggml.o utils.o -o chat $(LDFLAGS)
 
+chat_mac: chat.cpp ggml.c utils.cpp
+	$(CC)  $(CFLAGS)   -c ggml.c -o ggml_x86.o -target x86_64-apple-macos
+	$(CC)  $(CFLAGS)   -c ggml.c -o ggml_arm.o -target arm64-apple-macos
+	
+	$(CXX) $(CXXFLAGS) chat.cpp ggml_x86.o utils.cpp -o chat_x86 $(LDFLAGS) -target x86_64-apple-macos
+	$(CXX) $(CXXFLAGS) chat.cpp ggml_arm.o utils.cpp -o chat_arm $(LDFLAGS) -target arm64-apple-macos
+	lipo -create -output chat_mac chat_x86 chat_arm
 
 quantize: quantize.cpp ggml.o utils.o
 	$(CXX) $(CXXFLAGS) quantize.cpp ggml.o utils.o -o quantize $(LDFLAGS)

From 7636ddb9e45807ce4317717f75a4d7ce5d23541a Mon Sep 17 00:00:00 2001
From: Kevin Kwok <antimatter15@gmail.com>
Date: Mon, 20 Mar 2023 22:16:27 -0700
Subject: [PATCH 26/26] Update README.md

---
 README.md | 81 +++++++++++++++++++++++++++++++++----------------------
 1 file changed, 49 insertions(+), 32 deletions(-)

diff --git a/README.md b/README.md
index a1ecbe78cef8..aec100114a5a 100644
--- a/README.md
+++ b/README.md
@@ -8,37 +8,69 @@ Run a fast ChatGPT-like model locally on your device. The screencast below is no
 
 This combines the [LLaMA foundation model](https://github.com/facebookresearch/llama) with an [open reproduction](https://github.com/tloen/alpaca-lora) of [Stanford Alpaca](https://github.com/tatsu-lab/stanford_alpaca) a fine-tuning of the base model to obey instructions (akin to the [RLHF](https://huggingface.co/blog/rlhf) used to train ChatGPT) and a set of modifications to [llama.cpp](https://github.com/ggerganov/llama.cpp) to add a chat interface. 
 
-## Get started
+## Get Started (7B)
 
-```sh
-git clone https://github.com/antimatter15/alpaca.cpp
-cd alpaca.cpp
+Download the zip file corresponding to your operating system from the [latest release](https://github.com/antimatter15/alpaca.cpp/releases/latest). On Windows, download `alpaca-win.zip`, on Mac (both Intel or ARM) download `alpaca-mac.zip`, and on Linux (x64) download `alpaca-linux.zip`. 
 
-make chat
+Download `ggml-alpaca-7b-q4.bin` and place it in the same folder as the `chat` executable in the zip file. There are several options: 
+
+```
+# Any of these commands will work. 
+curl -o ggml-alpaca-7b-q4.bin -C - https://gateway.estuary.tech/gw/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
+curl -o ggml-alpaca-7b-q4.bin -C - https://ipfs.io/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
+curl -o ggml-alpaca-7b-q4.bin -C - https://cloudflare-ipfs.com/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
+
+# BitTorrent
+magnet:?xt=urn:btih:5aaceaec63b03e51a98f04fd5c42320b2a033010&dn=ggml-alpaca-7b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce
+https://btcache.me/torrent/5AACEAEC63B03E51A98F04FD5C42320B2A033010  
+https://torrage.info/torrent.php?h=5aaceaec63b03e51a98f04fd5c42320b2a033010  
+```
+
+Once you've downloaded the model weights and placed them into the same directory as the `chat` or `chat.exe` executable, run:
+
+```
 ./chat
 ```
 
-You can download the weights for `ggml-alpaca-7b-q4.bin` with BitTorrent:
+The weights are based on the published fine-tunes from `alpaca-lora`, converted back into a pytorch checkpoint with a [modified script](https://github.com/tloen/alpaca-lora/pull/19) and then quantized with llama.cpp the regular way. 
+
+## Getting Started (13B)
 
-magnet: `magnet:?xt=urn:btih:5aaceaec63b03e51a98f04fd5c42320b2a033010&dn=ggml-alpaca-7b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce`  
-torrent: https://btcache.me/torrent/5AACEAEC63B03E51A98F04FD5C42320B2A033010  
-torrent: https://torrage.info/torrent.php?h=5aaceaec63b03e51a98f04fd5c42320b2a033010  
+If you have more than 10GB of RAM, you can use the higher quality 13B `ggml-alpaca-13b-q4.bin` model. To download the weights, you can use
 
+```
 
-Alternatively you can download them with IPFS.
+# Any of these commands will work. 
+curl -o ggml-alpaca-13b-q4.bin -C - https://gateway.estuary.tech/gw/ipfs/Qme6wyw9MzqbrUMpFNVq42rC1kSdko7MGT9CL7o1u9Cv9G 
+curl -o ggml-alpaca-13b-q4.bin -C - https://ipfs.io/ipfs/Qme6wyw9MzqbrUMpFNVq42rC1kSdko7MGT9CL7o1u9Cv9G 
+curl -o ggml-alpaca-13b-q4.bin -C - https://cloudflare-ipfs.com/ipfs/Qme6wyw9MzqbrUMpFNVq42rC1kSdko7MGT9CL7o1u9Cv9G 
 
+# BitTorrent
+magnet:?xt=urn:btih:053b3d54d2e77ff020ebddf51dad681f2a651071&dn=ggml-alpaca-13b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce&tr=udp%3A%2F%2Ftracker.openbittorrent.com%3A6969%2Fannounce&tr=udp%3A%2F%2F9.rarbg.com%3A2810%2Fannounce
+https://btcache.me/torrent/053B3D54D2E77FF020EBDDF51DAD681F2A651071  
+https://torrage.info/torrent.php?h=053b3d54d2e77ff020ebddf51dad681f2a651071
 ```
-# any of these will work
-curl -o ggml-alpaca-7b-q4.bin -C - https://gateway.estuary.tech/gw/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
-curl -o ggml-alpaca-7b-q4.bin -C - https://ipfs.io/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
-curl -o ggml-alpaca-7b-q4.bin -C - https://cloudflare-ipfs.com/ipfs/QmQ1bf2BTnYxq73MFJWu1B7bQ2UD6qG7D7YDCxhTndVkPC
+
+Once you've downloaded the weights, you can run the following command to enter chat
+
+```
+./chat -m ggml-alpaca-13b-q4.bin
 ```
 
-Save the `ggml-alpaca-7b-q4.bin` file in the same directory as your `./chat` executable. 
 
-The weights are based on the published fine-tunes from `alpaca-lora`, converted back into a pytorch checkpoint with a [modified script](https://github.com/tloen/alpaca-lora/pull/19) and then quantized with llama.cpp the regular way. 
+## Building from Source (MacOS/Linux)
+
+
+```sh
+git clone https://github.com/antimatter15/alpaca.cpp
+cd alpaca.cpp
+
+make chat
+./chat
+```
+
 
-## Windows Setup
+## Building from Source (Windows)
 
 - Download and install CMake: <https://cmake.org/download/>
 - Download and install `git`. If you've never used git before, consider a GUI client like <https://desktop.github.com/>
@@ -59,21 +91,6 @@ cmake --build . --config Release
 - (You can add other launch options like `--n 8` as preferred onto the same line)
 - You can now type to the AI in the terminal and it will reply. Enjoy!
 
-## 13B
-
-TODO: write more docs here (PRs welcome)
-
-You can download the weights for `ggml-alpaca-13b-q4.bin` with BitTorrent:
-
-magnet: `magnet:?xt=urn:btih:053b3d54d2e77ff020ebddf51dad681f2a651071&dn=ggml-alpaca-13b-q4.bin&tr=udp%3A%2F%2Ftracker.opentrackr.org%3A1337%2Fannounce&tr=udp%3A%2F%2Fopentracker.i2p.rocks%3A6969%2Fannounce&tr=udp%3A%2F%2Ftracker.openbittorrent.com%3A6969%2Fannounce&tr=udp%3A%2F%2F9.rarbg.com%3A2810%2Fannounce`  
-torrent: https://btcache.me/torrent/053B3D54D2E77FF020EBDDF51DAD681F2A651071  
-torrent: https://torrage.info/torrent.php?h=053b3d54d2e77ff020ebddf51dad681f2a651071
-
-
-```
-./chat -m ggml-alpaca-13b-q4.bin
-```
-
 ## Credit
 
 This combines [Facebook's LLaMA](https://github.com/facebookresearch/llama), [Stanford Alpaca](https://crfm.stanford.edu/2023/03/13/alpaca.html), [alpaca-lora](https://github.com/tloen/alpaca-lora) and [corresponding weights](https://huggingface.co/tloen/alpaca-lora-7b/tree/main) by Eric Wang (which uses [Jason Phang's implementation of LLaMA](https://github.com/huggingface/transformers/pull/21955) on top of Hugging Face Transformers), and [llama.cpp](https://github.com/ggerganov/llama.cpp) by Georgi Gerganov. The chat implementation is based on Matvey Soloviev's [Interactive Mode](https://github.com/ggerganov/llama.cpp/pull/61) for llama.cpp. Inspired by [Simon Willison's](https://til.simonwillison.net/llms/llama-7b-m2) getting started guide for LLaMA. [Andy Matuschak](https://twitter.com/andy_matuschak/status/1636769182066053120)'s thread on adapting this to 13B, using fine tuning weights by [Sam Witteveen](https://huggingface.co/samwit/alpaca13B-lora).